Vous êtes sur la page 1sur 376


9781623568177_txt_print.indd 1 29/04/2013 11:50
Founding Editor
Francisco J. Ricardo
Associate Editor
Jrgen Schfer
Editorial Board
Roberto Simanowski
Rita Raley
John Cayley
George Field
9781623568177_txt_print.indd 2 29/04/2013 11:50
9781623568177_txt_print.indd 3 29/04/2013 11:50
Bloomsbury Academic
An imprint of Bloomsbury Publishing Plc
1385 Broadway 50 Bedford Square
New York London
NY 10018 WC1B 3DP
First published 2013
Lev Manovich, 2013
This work is published subject to a Creative Commons Attribution
Non-Commercial Licence.
You may share this work for non-commercial purposes only, provided you give
attribution to the copyright holder and the publisher. For permission to publish
commercial versions please contact Bloomsbury Academic.
Library of Congress Cataloging-in-Publication Data
Manovich, Lev.
Software takes command : extending the language of new media / by Lev
pages cm
Includes bibliographical references and index.
ISBN 978-1-62356-745-3 (pbk. : alk. paper) -- ISBN 978-1-62356-817-7 (hardcover :
alk. paper) 1. Multimedia systems. 2. Mass media--Technological innovations.
3. Computer graphics. I. Title.
QA76.575.M3157 2013
ISBN: HB: 978-1-6235-6817-7
PB: 978-1-6235-6745-3
e-pdf: 978-1-6235-6672-2
e-pub: 978-1-6235-6261-8
Typeset by Fakenham Prepress Solutions, Fakenham, Norfolk NR21 8NN
Printed and bound in the United States of America
9781623568177_txt_print.indd 4 29/04/2013 11:50
To Hyunjoo
9781623568177_txt_print.indd 5 29/04/2013 11:50
9781623568177_txt_print.indd 6 29/04/2013 11:50
Acknowledgments ix
Introduction 1
Understanding media 1
Software, or the engine of contemporary societies 6
What is software studies? 10
Cultural software 20
Media applications 24
From documents to performances 33
Why the history of cultural software does not exist 39
Summary of the books narrative 43
PART Inventing media software 53
1 Alan Kays universal media machine 55
Appearance versus function 55
Simulation is the central notion of the Dynabook 64
The permanent extendibility 91
The computer as a metamedium 101
2 Understanding metamedia 107
The building blocks 107
Media-independent vs. media-specic techniques 113
Inside Photoshop 124
There is only software 147
9781623568177_txt_print.indd 7 29/04/2013 11:50
PART Hybridization and evolution 159
3 Hybridization 161
Hybridity vs. multimedia 161
The evolution of a computer metamedium 176
Hybridity: examples 184
Strategies of hybridization 195
4 Soft evolution 199
Algorithms and data structures 199
What is a medium? 204
The metamedium or the monomedium? 225
The evolution of media species 233
PART Software in action 241
5 Media design 243
After Efects and the invisible revolution 243
The aesthetics of hybridity 254
Deep remixability 267
Layers, transparency, compositing 277
After Efects interface: from time-based to
composition-based 282
3D space as a media design platform 289
Import/export: design workow 296
Variable form 307
Amplication 323
Conclusion 329
Software, hardware, and social media 329
Media after software 335
Software epistemology 337
Index 343
9781623568177_txt_print.indd 8 29/04/2013 11:50
The ideas and arguments in this book are the result of the authors
interactions with hundreds of people over many years: students in
classes, presenters at conferences, colleagues over email. I especially
want to thank everybody who responded to my posts related to the
ideas in this book on Twitter and Facebook as I was working on
it from 2007 to 2012. They asked provocative questions, told me
about relevant resources every time I asked, and encouraged me to
go forward by asking when the book will be published.
The following people at a number of institutions played particularly
key roles in the books evolution and publication, and I would like to
thank them individually (they are listed alphabetically by institution):
Bloomsbury Academic (the book publisher):
Katie Gallof, Acquisitions Editor, Film and Media Studies.
Jennifer Laing, Copy-editing.
Francisco J. Ricardo, Editor, International Texts in Critical
Media Aesthetics.
Clare Turner, Lead Designer.
Software Studies Initiative (my lab established in 2007 at
Jeremy Douglass, Post-doctoral researcher, 20082012 (now
Assistant Professor, English Department, UCSB).
Jay Chow, stuf member, 2012 (design, visualization, and
Benjamin Bratton, Associate Professor, Visual Arts, UCSD.
Elisabeth Losh, Director of Academic Programs, Sixth
College, UCSD.
9781623568177_txt_print.indd 9 29/04/2013 11:50
California Institute for Telecommunication and Information
Hector Bracho, media services.
Doug Ramsey, Director of Communications.
Ramesh Rao, Director, UCSD Division, Calit2.
Larry Smarr, Director, Calit2.
Center for Research in Computing and the Arts (CRCA):
Sheldon Brown, Director; Professor, Visual Arts, UCSD.
Lourdes Guardiano-Durkin, MSO.
Todd Margolis, Technical Director.
The Graduate Center, City University of New York (CUNY):
Matthew Gold, Associate Professor; Director, CUNY
Academic Commons.
Tanya Domi, Director of Media Relations.
Chase Robinson, Provost and Senior Vice-Chancellor.
Jane Trombley, Executive Director for Communication and
Software Studies Series (The MIT Press):
Matthew Fuller and Noah Wardrip-Fruin (series
Douglass Sery, Editor, New Media, Game Studies, Design.
Finally, I want to add special thanks to Larry Smarr, Director
of California Institute for Telecommunication and Information
who invited me to participate in the Institute activities, helped to
start my lab, and made it possible for me to work with the next
generation of computing technologies being invented at Calit2 and
the people inventing them.
Large parts of the book were written and edited in my favorite
cafes and hotel lobbies, and I would like to thank their staf:
The Standard, West Hollywood, California.
Mondrian, West Hollywood, California.
LAuberge Del Mar, California.
Del Mar Plaza, Del Mar, California.
Starbucks, Del Mar, California.
Sheraton Tribeca Hotel, New York.
9781623568177_txt_print.indd 10 29/04/2013 11:50
The book was written on Apple laptops (MacBook Pro, MacBook
Air) using Microsoft Word. I used iPhone for email and social
networks, and for occasional note taking.
For communication with the colleagues and the publisher, I
relied on Gmail, Google Docs, Dropbox, Twitter, and Facebook.
The book illustrations were prepared by me and Jay Chow using
the same popular software applications analyzed in this book:
Photoshop, Illustrator, After Efects.
I am grateful to thousands of programmers and engineers who
developed the software products mentioned above, and continue
updating them with new features.
9781623568177_txt_print.indd 11 29/04/2013 11:50
9781623568177_txt_print.indd 12 29/04/2013 11:50
Understanding media
I called my earlier book-length account of the new cultural
forms enabled by computerization The Language of New Media
(completed in 1999, it came out in 2001). By that time, the process
of adoption of software-based tools in all areas of professional
media production was almost complete, and new media art was
in its heroic and vibrant stageofering many possibilities not yet
touched by commercial software and consumer electronics.
Ten years later, most media became new media. The devel-
opments of the 1990s have been disseminated to the hundreds of
millions of people who are writing blogs, uploading photos and
videos to media sharing sites, and use free media authoring and
editing software tools that ten years earlier would have cost tens of
thousands of dollars.
Thanks to the practices pioneered by Google, the world is now
used to running on web applications and services that have never
been ofcially completed but remain forever in Beta stage. Since
these applications and services run on the remote servers, they can
be updated anytime without consumers having to do anything
and in fact, Google is updating its search algorithm code a few
times a day. Similarly, Facebook is also updating its code daily,
and sometimes it breaks. (Facebooks motto expressed in posters
around its ofces is Move Fast and Break Things.) Welcome to
the world of permanent changethe world that is now dened
9781623568177_txt_print.indd 1 29/04/2013 11:50
not by heavy industrial machines that change infrequently, but by
software that is always in ux.
Why should humanists, social scientists, media scholars, and
cultural critics care about software? Because outside of certain
cultural areas such as crafts and ne art, software has replaced a
diverse array of physical, mechanical, and electronic technologies
used before the twenty-rst century to create, store, distribute and
access cultural artifacts. When you write a letter in Word (or its
open source alternative), you are using software. When you are
composing a blog post in Blogger or WordPress, you are using
software. When you tweet, post messages on Facebook, search
through billions of videos on YouTube, or read texts on Scribd,
you are using software (specically, its category referred to as web
applications or webwaresoftware which is accessed via web
browsers and which resides on the servers).
And when you play a video game, explore an interactive instal-
lation in a museum, design a building, create special efects for a
feature lm, design a website, use a mobile phone to read a movie
review or to view the actual movie, and carry out thousands of
other cultural activities, in practical terms, you are doing the same
thingusing software. Software has become our interface to the
world, to others, to our memory and our imaginationa universal
language through which the world speaks, and a universal engine
on which the world runs. What electricity and the combustion
engine were to the early twentieth century, software is to the early
twenty-rst century.
This book is concerned with media softwareprograms
such as Word, PowerPoint, Photoshop, Illustrator, After Efects,
Final Cut, Firefox, Blogger, WordPress, Google Earth, Maya, and
3ds Max. These programs enable creation, publishing, sharing,
and remixing of images, moving image sequences, 3D designs,
texts, maps, and interactive elements, as well as various combina-
tions of these elements such as websites, interactive applications,
motion graphics, virtual globes, and so on. Media software also
includes web browsers such as Firefox and Chrome, email and chat
programs, news readers, and other types of software applications
whose primary focus is accessing media content (although they
sometimes also include some authoring and editing features.)
These software tools for creating, interacting with, and sharing
media represent a particular subset of application software
9781623568177_txt_print.indd 2 29/04/2013 11:50
A digital studio in Seoul, South Korea, 1/2006. This small studio was
responsible for the photography of all Samsung phones, to appear in
its ads worldwide. In the photos we see studio staf adjusting the phone
photos in Photoshop. Later these high-resolution retouched images were
inserted in the Samsung TV ad, thus assuring that the product details are
clearly visible.
9781623568177_txt_print.indd 3 29/04/2013 11:50
(including web applications) in general. Given this, we may expect
that all these tools inherit certain traits common to all contem-
porary software. Does this mean that regardless of whether you are
working on designing a space, creating special efects for a feature
lm, designing a website, or making information graphics, your
design process may follow a similar logic? Are there some struc-
tural features which motion graphics, graphic designs, websites,
product designs, buildings, and video games share since they are
all designed with software? More generally, how are interfaces and
the tools of media authoring software shaping the contemporary
aesthetics and visual languages of diferent media forms?
Behind these questions investigated in this book lies another
theoretical question. This question drives the book narrative
and motivates my choice of topics. What happens to the idea
of a medium after previously media-specic tools have been
simulated and extended in software? Is it still meaningful to talk
about diferent mediums at all? Or do we now nd ourselves in a
new brave world of one single monomedium, or a metamedium (to
borrow the term of the books key protagonist Alan Kay)?
In short: What is media after software?
Does media still exist?
This book is a theoretical account of media software and its efects
on the practice and the very concept of media. Over the last two
decades, software has replaced most other media technologies
that emerged in the nineteenth and twentieth centuries. Today it
is ubiquitous and taken for grantedand yet, surprisingly, few
people know about its history and the theoretical ideas behind its
development. You are likely to know the names of Renaissance
artists who popularized the use of linear perspective in western art
(Brunelleschi, Alberti) or early twentieth-century inventors of modern
lm language (D. W. Grifth, Eisenstein, etc.)but I bet you do not
know where Photoshop comes from, or Word, or any other media
tool you are using every day. More importantly, you probably do
not know why these tools were invented in the rst place.
What is the intellectual history of media software? What was
the thinking and motivation of the key people and research groups
they were directingJ. C. R. Licklider, Ivan Sutherland, Ted
9781623568177_txt_print.indd 4 29/04/2013 11:50
Nelson, Douglas Engelbart, Alan Kay, Nicholas Negropontewho
between 1960 and the late 1970s created most of the concepts
and practical techniques that underlie todays media applications?
As I discoveredand I hope you will share my original surprise,
in reading my analysis of the original texts by these peoplethey
were as much media theoreticians as computer engineers. I will
discuss their media theories and test them in view of the digital
media developments in the subsequent decades. As we will see,
the theoretical ideas of these people and their collaborators work
very well today, helping us to better understand the contemporary
software we use to create, read, view, remix, and share.
Welcome, then, to the secret history of our software culture
secret not because it was deliberately hidden but because until
recently, excited by all the rapid transformations cultural comput-
erization was bringing about, we did not bother to examine its
origins. This book will try to convince you that such an exami-
nation is very much worth your time.
Its title pays homage to a seminal twentieth-century book
Mechanization Takes Command: a Contribution to Anonymous
History (1947) by architectural historian and critic Sigfried Giedion.
In this work Giedion traces the development of mechanization in
industrial society across a number of domains, including systems of
hygiene and waste management, fashion, agricultural production,
and food system, with separate sections of the book devoted to
bread, meat, and refrigeration. Much more modest in scope, my
book presents episodes from the history of softwarization (my
neologism) of culture between 1960 and 2010, with a particular
attention to media softwarefrom the original ideas which led to
its development to its current ubiquity.
My investigation is situated within a broader intellectual
paradigm of software studies. From this perspective, this books
contribution is the analysis of the ideas that eventually led to media
software, and the efects of the adoption of this type of software on
contemporary media design and visual culture.
Note that the category media software is a subset of the
category application software; this category in its turn is a
subset of the category software
which I understand to include
http://en.wikipedia.org/wiki/List_of_software_categories (July 7, 2011).
9781623568177_txt_print.indd 5 29/04/2013 11:50
not only application software, system software, and computer
programming tools, but also social network services and social
media technologies.
If we understand software in this extended sense, we can ask,
What does it mean to live in software society? And what does it
mean to be part of software culture? These are the questions the
next section will take up.
Software, or the engine of
contemporary societies
In the beginning of the 1990s, the most famous global brands were
the companies that were in the business of producing materials or
goods, or processing physical matter. Today, however, the lists of
best-recognized global brands are topped with the names such as
Google, Facebook, and Microsoft. (In fact, in 2007 Google became
number one in the world in terms of brand recognition.) And, at
least in the US, the most widely read newspapers and magazines
New York Times, USA Today, Business Week, etc.feature daily
news and stories about Facebook, Twitter, Apple, Google, and
other IT companies.
What about other media? When I was working on the rst
draft of this book in 2008, I checked the business section of the
CNN website. Its landing page displayed market data for just ten
companies and indexes.
Although the list was changed daily, it
was always likely to include some of the same IT brands. Let us
take January 21, 2008 as an example. On that day the CNN list
contained the following companies and indexes: Google, Apple,
S&P 500 Index, Nasdaq Composite Index, Dow Jones Industrial
Andreas Kaplan and Michael Haenlein dene social media as a group of Internet-
based applications that build on the ideological and technological foundations of
Web 2.0, which allows the creation and exchange of user-generated content.
Andreas Kaplan and Michael Haenlein, Users of the world, unite! The challenges
and opportunities of Social Media, Business Horizons 53, no. 1 (JanuaryFebruary
2010), pp. 5968, http://dx.doi.org/10.1016/j.bushor.2009.09.003
http://money.cnn.com (January 21, 2008).
9781623568177_txt_print.indd 6 29/04/2013 11:50
Average, Cisco Systems, General Electric, General Motors, Ford,

This list is very telling. The companies that deal with physical
goods and energy appear in the second part of the list: General
Electric, General Motors, Ford. Right before and after these three,
we see two IT companies that provide hardware: Intel makes
computer chips, while Cisco makes network equipment. What about
the two companies which are on top: Google and Apple? The rst
is in the business of information (Googles mission is to organize
the worlds information and make it universally accessible and
), while the second is making consumer electronics: phones,
tablets, laptops, monitors, music players, etc. But actually, they are
both making something else. And apparently, this something else
is so crucial to the workings of US economyand consequently,
global world as wellthat these companies almost daily appear in
business news. And the major Internet companies that also appear
daily in news such as Google, Facebook, Twitter, Amazon, eBay,
and Yahoo, are in the same business.
This something else is software. Search engines, recommen-
dation systems, mapping applications, blog tools, auction tools,
instant messaging clients, and, of course, platforms which allow
people to write new softwareiOS, Android, Facebook, Windows,
Linuxare in the center of the global economy, culture, social life,
and, increasingly, politics. And this cultural softwarecultural
in a sense that it is directly used by hundreds of millions of people
and that it carries atoms of cultureis only the visible part of a
much larger software universe.
In Software Society (2003), an unrealized book proposal put
together by me and Benjamin Bratton, we described the importance
of software and its relative invisibility in humanities and social
science research:
Software controls the ight of a smart missile toward its target
during war, adjusting its course throughout the ight. Software
runs the warehouses and production lines of Amazon, Gap,
Dell, and numerous other companies allowing them to assemble
and dispatch material objects around the world, almost in no
http://www.google.com/about/company/ (September 23, 2012).
9781623568177_txt_print.indd 7 29/04/2013 11:50
time. Software allows shops and supermarkets to automati-
cally restock their shelves, as well as automatically determine
which items should go on sale, for how much, and when and
where in the store. Software, of course, is what organizes the
Internet, routing email messages, delivering Web pages from
a server, switching network trafc, assigning IP addresses, and
rendering Web pages in a browser. The school and the hospital,
the military base and the scientic laboratory, the airport and
the cityall social, economic, and cultural systems of modern
societyrun on software. Software is the invisible glue that ties
it all together. While various systems of modern society speak in
diferent languages and have diferent goals, they all share the
syntaxes of software: control statements if then and while
do, operators and data types (such as characters and oating
point numbers), data structures such as lists, and interface
conventions encompassing menus and dialog boxes.
If electricity and the combustion engine made industrial
society possible, software similarly enables global information
society. The knowledge workers, the symbol analysts, the
creative industries, and the service industriesnone of
these key economic players of the information society can exist
without software.
Examples are data visualization software used by a scientist,
spreadsheet software used by a nancial analyst, Web design
software used by a designer working for a transnational adver-
tising agency, or reservation software used by an airline. Software
is what also drives the process of globalization, allowing
companies to distribute management nodes, production facil-
ities, and storage and consumption outputs around the world.
Regardless of which new dimension of contemporary existence a
particular social theory of the last few decades has focused on
information society, knowledge society, or network societyall
these new dimensions are enabled by software.
Paradoxically, while social scientists, philosophers, cultural
critics, and media and new media theorists seem by now to cover
all aspects of IT revolution, creating a number of new disciplines
such as cyberculture studies, Internet studies, game studies, new
media theory, digital culture, and digital humanities, the under-
lying engine which drives most of these subjectssoftwarehas
received comparatively little attention.
9781623568177_txt_print.indd 8 29/04/2013 11:50
Even today, ten years later, when people are constantly interacting
with and updating dozens of apps on their mobile phones and
other computer devices, software as a theoretical category is still
invisible to most academics, artists, and cultural professionals
interested in IT and its cultural and social efects.
There are some important exceptions. One is the open source
movement and related issues around copyright and IP that have
been extensively discussed in many academic disciplines. We also
see a steadily growing number of trade books about Google,
Facebook, Amazon, eBay, Oracle, and other web giants. Some of
these books ofer insightful discussions of the software developed
by these companies and the social, political, cognitive, and episte-
mological efects of this software. (For a good example, see John
Battelle, The Search: How Google and Its Rivals Rewrote the Rules
of Business and Transformed Our Culture.
So while we are in a better situation today when we put together
our proposal for Software Society in 2003, I feel that it is still
meaningful to quote it (the only additions are the references to
social media and crowdsourcing):
If we limit critical discussions of digital culture to the notions
of open access, peer production, cyber, digital,
Internet, networks, new media, or social media, we
will never be able to get to what is behind new representational
and communication media and to understand what it really is
and what it does. If we dont address software itself, we are in
danger of always dealing only with its efects rather than the
causes: the output that appears on a computer screen rather than
the programs and social cultures that produce these outputs.
Information society, knowledge society, network society,
social media, online collaboration, crowdsourcing
regardless of which new feature of contemporary existence a
particular analysis has focused on, all these new features are
enabled by software. It is time we focused on software itself.
A similar sentiment is expressed in Noah Wardrip-Fruins
Expressive Processing (2009) when he says in relation to books
John Battelle, The Search: How Google and Its Rivals Rewrote the Rules of
Business and Transformed Our Culture (Portfolio Trade, 2006).
9781623568177_txt_print.indd 9 29/04/2013 11:50
about digital literature: almost all of these have focused on
what the machines of digital media look like from the outside:
their output regardless of perspective, writings on digital media
almost all ignore something crucial: the actual processes that make
digital media work, the computational machines that make digital
media possible.
My book discusses what I take to be the key part
of these machines today (because it is the only part which most
users see and use directly): application software.
What is software studies?
This book aims to contribute to the developing intellectual
paradigm of software studies. What is software studies? Here
are a few denitions. The rst comes from my The Language
of New Media, where, as far as I know, the terms software
studies and software theory appeared for the rst time. I
wrote, New media calls for a new stage in media theory whose
beginnings can be traced back to the revolutionary works of
Robert Innis and Marshall McLuhan of the 1950s. To under-
stand the logic of new media we need to turn to computer
science. It is there that we may expect to nd the new terms,
categories, and operations that characterize media that became
programmable. From media studies, we move to something
which can be called software studies; from media theoryto
software theory.
Reading this statement today, I feel some adjustments are in
order. It positions computer science as a kind of absolute truth,
a given which can explain to us how culture works in software
society. But computer science is itself part of culture. Therefore, I
think that Software Studies has to investigate the role of software
in contemporary culture, and the cultural and social forces that are
shaping the development of software itself.
The book that rst comprehensively demonstrated the necessity
of the second approach was New Media Reader edited by Noah
Wardrip-Fruin and Nick Montfort (The MIT Press, 2003). The
publication of this groundbreaking anthology laid the framework
Noah Wardrip-Fruin, Expressive Processing (Cambridge, MA: The MIT Press, 2009).
9781623568177_txt_print.indd 10 29/04/2013 11:50
for the historical study of software as it relates to the history
of culture. Although Reader did not explicitly use the term
software studies, it did propose a new model for how to think
about software. By systematically juxtaposing important texts by
pioneers of cultural computing and artists and writers active in
the same historical periods, New Media Reader demonstrated that
both belonged to the same larger epistemes. That is, often the same
idea was simultaneously articulated independently by artists and
scientists who were inventing cultural computing. For instance, the
anthology opens with a story by Jorge Borges (1941) and an article
by Vannevar Bush (1945) which both contain the idea of a massive
branching structure as a better way to organize data and to capture
human experience.
In February 2006 Matthew Fuller who had already published
a pioneering book on software as culture (Behind the Blip:
essays on the culture of software, 2003) organized the very rst
Software Studies Workshop at Piet Zwart Institute in Rotterdam.
Introducing the workshop, Fuller wrote, Software is often a blind
spot in the theorization and study of computational and networked
digital media. It is the very grounds and stuf of media design. In a
sense, all intellectual work is now software study, in that software
provides its media and its context, but there are very few places
where the specic nature, the materiality, of software is studied
except as a matter of engineering.

I completely agree with Fuller that, all intellectual work is
now software study. Yet it will take some time before the intel-
lectuals will realize it. To help bring this change, in 2008, Matthew
Fuller, Noah Wardrip-Fruin and I established the Software Studies
book series at MIT Press. The already published books in the
series are Software Studies: A Lexicon edited by Fuller (2008),
Expressive Processing: Digital Fictions, Computer Games, and
Software Studies by Wardrip-Fruin (2009), Programmed Visions:
Software and Memory by Wendy Hui Kyong Chun (2011), Code/
Space: Software and Everyday Life by Rob Kitchin and Martin
Dodge (2011), and Speaking Code: Coding as Aesthetic and
Political Expression by Geof Cox and Alex Mclean (2012). In
2011, Fuller together with a number of UK researchers established
http://pzwart.wdka.hro.nl/mdr/Seminars2/softstudworkshop (January 21, 2008).
9781623568177_txt_print.indd 11 29/04/2013 11:50
Computational Culture, an open-access peer-reviewed journal that
provides a platform for more publications and discussions.
In addition to this series, I am also happy to see a growing
number of other titles written from the perspectives of platform
studies, digital humanities, cyberculture, internet studies, and
game studies. Many of these books contain important insights
and discussions which help us better understand the roles of
software. Rather than trying to list all of them, I will only
provide a few examples of works which exemplify the rst two
of these perspectives (more will be in press by the time you are
reading this). Platform studies: Nick Montfort and Ian Bogosts
Racing the Beam: The Atari Video Computer System (2009),
Jimmy Mahers The Future Was Here: The Commodore Amiga
(2012). Digital Humanities: Mechanisms: New Media and the
Forensic Imagination (Matthew G. Kirschenbaum, 2008), The
Philosophy of Software: Code and Mediation in the Digital Age
(David Berry, 2011), Reading Machines: Toward an Algorithmic
Criticism (Stephen Ramsay, 2011), How We Think: Digital Media
and Contemporary Technogenesis (Katherine Hayles, 2012).
highly relevant is the rst book in what may become a new area
of format studies: MP3: The Meaning of a Format (Jonathan
Sterne, 2012).
Another set of works which are relevant to understanding the
roles and functioning of software systems comes from people who
were trained in computer science but are also equally at home in
cultural theory, philosophy, digital art, or other humanistic elds:
Phoebe Sengers, Warren Sack, Fox Harrell, Michael Mateas, Paul
Dourish, and Phil Agre.
Yet another relevant category of books comprises the historical
studies of important labs and research groups central to the
development of modern software, other key parts of information
technology such as the internet, and professional practices of
Nick Montfort and Ian Bogost, Racing the Beam: The Atari Video Computer
System (The MIT Press, 2009); Jimmy Maher, The Future Was Here: The
Commodore Amiga (The MIT Press, 2012); David Berry, The Philosophy of
Software: Code and Mediation in the Digital Age (Palgrave Macmillan, 2011);
Stephen Ramsay, Reading Machines: Toward an Algorithmic Criticism (University
of Illinois Press, 2011), Katherine Hayles, How We Think: Digital Media and
Contemporary Technogenesis (University of Chicago Press, 2012).
Jonathan Sterne, MP3: The Meaning of a Format (Duke University Press, 2012).
9781623568177_txt_print.indd 12 29/04/2013 11:50
software engineering such as user testing. The examples of these
works listed chronologically are Katie Hafner and Matthew Lyons
Where Wizards Stay Up Late: The Origins Of The Internet
(1998), Michael Hiltziks Dealers of Lightning: Xerox PARC and
the Dawn of the Computer Age (2000), Martin Campbell-Kellys
From Airline Reservations to Sonic the Hedgehog: A History of the
Software Industry (2004), and Nathan Ensmengers The Computer
Boys Take Over: Computers, Programmers, and the Politics of
Technical Expertise (2010).

My all-time favorite book, however, remains Tools for Thought
published by Howard Rheingold in 1985, right at the moment when
domestication of computers and software starts, eventually leading
to their current ubiquity. This book is organized around the key
insight that computers and software are not just technology but
rather the new medium in which we can think and imagine difer-
ently. Similar understanding was shared by all the heroes of this
book who, with their collaborators, invented the computational
tools for thoughtsJ. C. R. Licklider, Ted Nelson, Douglass
Engelbart, Bob Taylor, Alan Kay, Nicholas Negroponte. (Today
many academics in humanities and social sciences still need to
grasp this simple but fundamental idea. They continue to think of
software as being strictly the domain of the Academic Computing
Department in their universitiessomething which is only there to
help them become more efcient, as opposed to the medium where
human intellectual creativity now dwells.)
This short sketch of the intellectual landscape around software
studies will be very incomplete if I do not mention the role of
artists in pioneering the cultural discussions of software. Beginning
around 2000, a number of artists and writers started to develop
the practice of software art which included exhibitions, festivals,
publishing books, and organizing online repositories of relevant
works. The key gures in these developments were Amy Alexander,
Katie Hafner and Matthew Lyon, Where Wizards Stay Up Late: The Origins Of
The Internet (Simon & Schuster, 1998); Michael A. Hiltzik, Dealers of Lightning:
Xerox PARC and the Dawn of the Computer Age (HarperBusiness, 2000); Martin
Campbell-Kelly, From Airline Reservations to Sonic the Hedgehog: A History of
the Software Industry (The MIT Press, 2004); Nathan L. Ensmenger, The Computer
Boys Take Over: Computers, Programmers, and the Politics of Technical Expertise
(The MIT Press, 2010).
9781623568177_txt_print.indd 13 29/04/2013 11:50
Inke Arns, Adrian Ward, Geof Cox, Florian Cramer, Matthew
Fuller, Olga Goriunova, Alex McLean, Alessandro Ludovico, Pit
Schultz, and Alexei Shulgin. In 2002 Christiane Paul organized
CODeDOCan exhibition of artistic codeat The Whitney
Museum of American Art;
in 2003, the major festival of digital
art Ars Electronica choose Code as its topic; and since 2001,
the transmediale festival has included artistic software as one
of its categories, and devoted a signicant space to it in the festi-
vals symposiums. Some of the software art projects pioneered the
examination of code as the new cultural and social artifact; others
ofered critical commentary on commercial software practices. For
example, Adrian Ward created an ironic Auto-Illustratoran
experimental, semi-autonomous, generative software artwork and
a fully functional vector graphic design application to sit alongside
your existing professional graphic design utilities.
Recognizing that the bits of software studies exist across many
books and art projects, Fuller writes in the Foreword to The MIT
Press Software Studies book series:
Software is deeply woven into contemporary lifeeconomically,
culturally, creatively, politicallyin manners both obvious and
nearly invisible. Yet while much is written about how software
is used, and the activities that it supports and shapes, thinking
about software itself has remained largely technical for much of its
history. Increasingly, however, artists, scientists, engineers, hackers,
designers, and scholars in the humanities and social sciences are
nding that for the questions they face, and the things they need to
build, an expanded understanding of software is necessary. For such
understanding they can call upon a strand of texts in the history of
computing and new media, they can take part in the rich implicit
culture of software, and they also can take part in the development
of an emerging, fundamentally transdisciplinary, computational
literacy. These provide the foundation for Software Studies.
Indeed, a number of earlier works by the leading media theorists
of our timesFriedrich A. Kittler, Peter Weibel, Katherine Hayles,
Matthew Fuller, Software Studies series introduction, http://mitpress.mit.edu/
catalog/browse/browse.asp?btype=6&serid=179 (July 14, 2011).
9781623568177_txt_print.indd 14 29/04/2013 11:50
Lawrence Lessig, Manual Castells, Alex Galloway, and otherscan
also be retroactively identied as belonging to software studies.

Therefore, I strongly believe that this paradigm has already existed
for a number of years but it has not been explicitly named until a
few years ago.
In his introduction to a 2006 Rotterdam workshop Fuller
pointed out that software can be seen as an object of study and
an area of practice for art and design theory and the humanities,
for cultural studies and science and technology studies and for
an emerging reexive strand of computer science. Since a new
academic discipline can be dened either through a unique object
of study, a new research method, or a combination of the two,
how shall we think of software studies? Fullers statement implies
that software is a new object of study which should be put on
the agenda of existing disciplines and which can be studied using
already existing methodsfor instance, actor-network theory,
social semiotics, or media archaeology.
There are good reasons for supporting this perspective. I think
of software as a layer that permeates all areas of contemporary
societies. Therefore, if we want to understand contemporary
techniques of control, communication, representation, simulation,
analysis, decision-making, memory, vision, writing, and interaction,
our analysis cannot be complete until we consider this software
layer. Which means that all disciplines which deal with contem-
porary society and culturearchitecture, design, art criticism,
sociology, political science, art history, media studies, science and
technology studies, and all othersneed to account for the role of
software and its efects in whatever subjects they investigate.
At the same time, the existing work in software studies already
demonstrates that if we are to focus on software itself, we need
new methodologies. That is, it helps to practice what one writes
about. It is not accidental that all the intellectuals who have most
systematically written about softwares roles in society and culture
have either programmed themselves or have been involved in
cultural projects and practices which include writing and teaching
softwarefor instance, Ian Bogost, Jay Bolter, Florian Cramer,
Wendy Chun, Matthew Fuller, Alexander Galloway, Katherine
See Michael Truscello, review of Behind the Blip: Essays on the Culture of
Software, Cultural Critique 63, Spring 2006, pp. 1827.
9781623568177_txt_print.indd 15 29/04/2013 11:50
Hayles, Matthew Kirschenbaum, Geert Lovink, Peter Lunenfeld,
Adrian Mackenzie, Paul D. Miller, William J. Mitchell, Nick
Montfort, Janet Murray, Katie Salen, Bruce Sterling, Noah Wardrip-
Fruin, and Eric Zimmerman. In contrast, the scholars without
this technical experience or involvementfor example, Manual
Castells, Bruno Latour, Paul Virilio, and Siegfried Zielinskihave
not included discussions of software in their otherwise theoreti-
cally precise and highly inuential accounts of modern media and
In the 2000s, the number of students in media art, design, archi-
tecture, and humanities who use programming or scripting in their
work has grown substantiallyat least in comparison with 1999
when I rst mentioned software studies in The Language of New
Media. Outside of culture and academic industries, many more
people today are also writing software. To a signicant extent, this
is the result of new programming and scripting languages such as
ActionScript, PHP, Perl, Python, and Processing. Another important
factor is the publication of APIs by all major Web 2.0 companies
in the middle of the 2000s. (API, or Application Programming
Interface, is a code that allows other computer programs to
access services ofered by an application. For instance, people can
use Google Maps API to embed full Google Maps on their own
websites.) These programming and scripting languages and APIs
did not necessarily make programming easier. Rather, they made
it much more efcient. For instance, since a young designer can
create an interesting work with only couple of dozen lines of code
written in Processing versus writing a really long Java program,
s/he is much more likely to take up programming. Similarly, if only
a few lines in JavaScript allows you to integrate all the functionality
ofered by Google Maps into your site, this is a great motivation
for beginning to work with JavaScript. Yet another reason for
more people writing software today is the emergence of a massive
mobile apps marketplace that, unlike the desktop market, is not
dominated by a few large companies. According to informal
reports in the beginning of 2012, one million programmers were
creating apps for the iOS platform (iPad and iPhone) alone, and
another one million were doing this for the Android platform.
In his 2006 article covering new technologies that allow people
with very little or no programming experience to create new
custom software (such as Ning), Martin LaMonica wrote about a
9781623568177_txt_print.indd 16 29/04/2013 11:50
future possibility of a long tail for apps.
A few years later, this
is exactly what happened. In September 2012, 700,000 apps were
available on Apple App Store,
and over 600,000 Android apps on
Google Play.

In the article called A Surge in Learning the Language of
the Internet (March 27, 2012), the New York Times reported
that, The market for night classes and online instruction in
programming and Web construction, as well as for iPhone apps
that teach, is booming. The article quoted Zach Sims, one
of the founders of Codecademy (a web school which teaches
programming though interactive lessons) who explained one of the
reasons for this growing interest in learning programming and web
design: People have a genuine desire to understand the world we
now live in. They do not just want to use the Web; they want to
understand how it works.
In spite of these impressive developments, the gap between
people who can program and who cannot remainsas does the
gap between professional programmers and people who just
took one or two short programming classes. Clearly, today the
consumer technologies for capturing and editing media are much
easier to use than even the most friendly programming and
scripting languages. But it does not necessarily have to stay this
way. Think, for instance, of what it took to set up a photo studio
and take photographs in the 1850s versus simply pressing a single
button on a digital camera or a mobile phone in the 2000s. Clearly,
we are very far from such simplicity in programming. But I do not
see any logical reasons why programming cannot one day become
equally easy.
For now, the number of people who can script and program
keeps increasing. Although we are far from a true long tail
for software, software development is gradually getting more
democratized. It is, therefore, the right moment to start thinking
Martin LaMonica, The do-it-yourself Web emerges, CNET News, July 31, 2006,
http://www.mobilestatistics.com/mobile-statistics (July 30, 2012).
http://play.google.com/about/apps/ (July 30, 2012).
Jenna Wortham, A Surge in Learning the Language of the Internet, New York
Times, March 27, 2012, http://www.nytimes.com/2012/03/28/technology/for-an-
9781623568177_txt_print.indd 17 29/04/2013 11:50
The complete code for tree_recursion, a Processing sketch by Mitchell
Whitelaw, 2011, http://www.openprocessing.org/sketch/8752
9781623568177_txt_print.indd 18 29/04/2013 11:50
Tree variations generated by tree_recursion code.
9781623568177_txt_print.indd 19 29/04/2013 11:50
theoretically about how software is shaping our culture, and how
it is shaped by culture in its turn. The time for software studies
has arrived.
Cultural software
German media and literary theorist Friedrich Kittler wrote that
students today should know at least two software languages:
only then theyll be able to say something about what culture
is at the moment.
Kittler himself programmed in an assembler
languagewhich probably determined his distrust of Graphical
User Interfaces and modern software applications that use these
interfaces. In a classical modernist move, Kittler argued that we
need to focus on the essence of the computerwhich for Kittler
meant its mathematical and logical foundations and its early
history characterized by tools such as assembler languages.
This book is determined by my own history of engagement with
computers as a programmer, computer animator and designer,
media artist, and as a teacher. This involvement started in the early
1980s, which was the decade of procedural programming (Pascal),
rather than assembly programming. It was also the decade that
saw the introduction of PCs, the emergence and popularization
of desktop publishing, and the use of hypertext by some literary
scholars. In fact, I came to NYC from Moscow in 1981, which
was the year IBM introduced their rst PC. My rst experience
with computer graphics was in 19834 on Apple IIe. In 1984 I
saw a Graphical User Interface in its rst successful commercial
implementation on an Apple Macintosh. The same year I got a job
at Digital Efects, one of the rst computer animation companies in
the world, where I learned how to program 3D computer models
and animations. In 1986 I was writing computer programs that
automatically processed photographs to make them look like
paintings. In January 1987 Adobe Systems shipped Illustrator,
Friedrich Kittler, Technologies of Writing/Rewriting Technology, Auseinander 1,
no. 3 (Berlin, 1995), quoted in Michael Truscello, The Birth of Software Studies:
Lev Manovich and Digital Materialism, Film-Philosophy 7, no. 55 (December
2003), http://www.film-philosophy.com/vol7-2003/n55truscello.html
9781623568177_txt_print.indd 20 29/04/2013 11:50
followed by Photoshop in 1989. The same year saw the release
of The Abyss, directed by James Cameron. This movie used
pioneering CGI to create the rst complex virtual character. And,
by Christmas of 1990, Tim Berners-Lee had already created all the
components of the World Wide Web as it exists today: a web server,
web pages, and a web browser.
In short, during one decade the computer moved from being a
culturally invisible technology to being the new engine of culture.
While the progress in hardware and Moores Law played crucial
roles in this development, even more crucial was the release
of software with a Graphical User Interface (GUI) aimed at
non-technical users, word processing, applications for drawing,
painting, 3D modeling, animation, music composing and editing,
information management, hypermedia and multimedia authoring
(HyperCard, Director), and global networking (World Wide Web)
With easy-to-use software in place, the stage was set for the next
decade of the 1990s when most culture industriesgraphic design,
architecture, product design, space design, lmmaking, animation,
media design, music, higher education, and culture management
gradually adapted software tools. Thus, although I rst learned to
program in 1975 when I was in high school in Moscow, my take
on software studies has been shaped by watching how during the
1980s GUI-based software quickly put the computer in the center
of culture.
If software is indeed the contemporary equivalent of the
combustion engine and electricity in terms of its social efects,
every type of software needs to be taken into account. We need to
consider not only visible software used by consumers but also
grey software, which runs all systems and processes in contem-
porary society. However, since I do not have personal experience
writing logistics software, industrial automation software, and
other grey software, I will be not be writing about such topics.
My concern is with a particular subset of software which I used
and taught in my professional life. I call it cultural software.
While the term cultural software was previously used
metaphorically (see J. M. Balkin, Cultural Software: A Theory of
Ideology, 2003), I am going to use it literally to refer to certain
types of software that support actions we normally associate
with culture. These cultural actions enabled by software can be
divided into a number of categories (of course we should keep in
9781623568177_txt_print.indd 21 29/04/2013 11:50
Adobe Photoshop, Macintosh version 1.0.7, 1990. Top: preferences
window. Bottom: workspace.
9781623568177_txt_print.indd 22 29/04/2013 11:50
mind that this is just one possible specic categorization system
among many).
1 Creating cultural artifacts and interactive services which
contain representations, ideas, beliefs, and aesthetic values
(for instance, editing a music video, designing a package for
a product, designing a website or an app).
2 Accessing, appending, sharing, and remixing such artifacts
(or their parts) online (for instance, reading newspaper on
the web, watching YouTube video, adding comments to a
blog post).
3 Creating and sharing information and knowledge online
(for instance, editing a Wikipedia article, adding places in
Google Earth, including a link in a tweet).
4 Communicating with other people using email, instant
message, voice-over IP, online text and video chat, social
networking features such as wall postings, pokes, events,
photo tags, notes, places, etc.
5 Engaging in interactive cultural experiences (for instance,
playing a computer game).
6 Participating in the online information ecology by
expressing preferences and adding metadata (for instance,
automatically generating new information for Google
Search whenever you use this service; clicking the +1
button on Google+ or the Like button on Facebook;
using the retweet function on Twitter).
7 Developing software tools and services that support all
these activities (for instance, programming a library for
Processing that enables sending and receiving data over the
writing a new plugin for Photoshop, creating a
new theme for WordPress).
Technically, this software may be implemented in a variety of ways.
Popular implementations (referred to in the computer industry as
software architecture) include stand-alone applications that run
on the users computing device, distributed applications (a client
http://www.processing.org/reference/libraries/ (July 7, 2011).
9781623568177_txt_print.indd 23 29/04/2013 11:50
running on the users device communicates with software on the
server), and peer-to-peer networks (each computer becomes both
a client and a server). If all this sounds completely unfamiliar, do
not worry: all you need to understand is that cultural software
as I will use this term covers a wide range of products and network
services, as opposed to only single desktop applications such
as Illustrator, Photoshop or After Efects that dominated media
authoring in the 1990s and 2000s. For example, social network
services such as Facebook and Twitter include multiple programs
and databases running on company servers (for instance, in 2007
Google was running over one million servers around the world
according to one estimate
) and the programs (called clients)
used by people to send emails, chat, post updates, upload video,
leave comments, and perform other tasks on these services. (For
instance, one can access Twitter using twitter.com, or tweetdeck.
com, Twitter apps for iOS, Android, and dozens of third party
websites and apps.)
Media applications
Let us go through the software categories that support the rst four
types of cultural activities listed above in more detail.
The rst category is software for creating, editing, and organizing
media content. The examples are Microsoft Word, PowerPoint,
Photoshop, Illustrator, InDesign, Final Cut, After Efects, Maya,
Blender, Dreamweaver, Aperture, and other applications. This
category is in the center of this book. The industry uses a number
of terms to refer to this category such as media authoring,
media editing, and media development but I am going to refer
to this category by using a single summary term. I will simply call
it media software.
The second category is software for distributing, accessing, and
combining (or publishing, sharing, and remixing) media
content on the web. Think Firefox, Chrome, Blogger, WordPress,
Tumblr, Pinterest, Gmail, Google Maps, YouTube, Vimeo and
Pandia Search & Social, Google: one million servers and counting, http://www.
9781623568177_txt_print.indd 24 29/04/2013 11:50
other web applications and services. Obviously, the rst and second
categories overlapfor example, many desktop media applications
allow you to upload your creations directly to popular media
sharing sites, while many web applications and services include
some authoring and editing functions (for example, YouTube has
a built-in video editor). And blogging platforms and email clients
sit right in the middlethey are used as much for publishing as for
creating new content.
I will take for granted that since we all use application programs,
or apps, we have a basic understanding of this term. Similarly, I
also assume that we understand what content refers to in digital
culture, but just to be sure, here are a couple of ways to dene
it. We can simply list various types of media which are created,
shared, and accessed with media software and the tools provided
by social media and sites: texts, images, digital video, animations,
3D objects and scenes, maps, as well as various combinations of
these and other media. Alternatively, we can dene content by
listing genres, for instance, web pages, tweets, Facebook updates,
casual games, multiplayer online games, user-generated video,
search engine results, URLs, map locations, shared bookmarks, etc.
Digital culture tends to modularize content, i.e., enabling users
to create, distribute, and re-use discrete content elementslooping
animations to be used as backgrounds for videos, 3D objects to
be used in creating complex 3D animations, pieces of code to be
used in websites and blogs, etc.
(This modularity parallels the
fundamental principle of modern software engineering to design
computer programs from small reusable parts called functions or
procedures.) All such parts also qualify as content.
Between the late 1970s and the middle of the 2000s, appli-
cation programs for media editing were designed to run on a
users computer (minicomputers, PCs, scientic workstations, and
later, laptops). In the next ve years, companies gradually created
more and more capable versions of these programs running in
the cloud. Some of these programs are available via their own
websites (Google Docs, Microsoft Web Ofce), while others are
integrated with media hosting or social media services (e.g.,
Photobucket image and video editor). Many applications are
implemented as clients that run on mobile phones (e.g., Maps
on iPhone), tablets, and TV platforms and communicate with
servers and websites. Examples of such platforms are Apples iOS,
9781623568177_txt_print.indd 25 29/04/2013 11:50
Googles Android, and LGs Smart TV App platform. Still others
are apps running on tablets such as Adobe Photoshop Touch for
(While at the moment of writing both web-based and
mobile applications have limited editing capabilities in comparison
with their desktop counterparts, this may already have changed by
the time you are reading this book).
The development of mobile software platforms led to the
increasing importance of certain media application types (and
corresponding cultural activities) such as media uploaders (apps
designed for uploading media content to media sharing sites). To
put this diferently, managing media content (for example, organ-
izing photos in Picasa) and also meta-managing (i.e. managing
the systems which manage it such as organizing a blogroll) have
become as central to a persons cultural life as creating this content.
This book is about media softwareits conceptual history, the
ways it redened the practice of media design, the aesthetics of
the media being created, and creators and users understanding of
media. How can we place media software inside other categories
and also break it into smaller categories? Let us start again
with our denition, which I will rephrase here. Media software
are programs that are used to create and interact with media
objects and environments. It is a subset of the larger category of
application softwarethe term which is itself in the process of
changing its meaning as desktop applications (applications which
run on a computer) are supplemented by mobile apps (applications
running on mobile devices) and web applications (applications
which consist of a web client and the software running on a
server). Media software enables creation, publishing, accessing,
sharing, and remixing diferent types of media (such as images,
moving image sequences, 3D shapes, characters, and spaces, text,
maps, interactive elements), as well as various projects and services
which use these elements. These projects can be non-interactive
(2D designs, motion graphics, lm shots) or interactive (media
surfaces and other interactive installations). The online services
are by their very nature always interactive (websites, blogs, social
networks, social media services, games, wikis, web media and app
stores such as Google Play and Apple iTunes, other shopping sites,
http://www.adobe.com/products/mobileapps/ (March 12, 2012).
9781623568177_txt_print.indd 26 29/04/2013 11:50
and so on)while a user is not always given the ability to add to
or modify content, s/he always navigates and interacts with the
existing content using interactive interface.
Given that today the multi-billion global culture industry is
enabled by media applications, it is interesting that there is no
single accepted way to classify them. The Wikipedia article on
application software includes the categories of media devel-
opment software and content access software (divided into web
browsers, media players, and presentation applications).
This is
generally useful but not completely accuratesince today most
content access software also includes at least some media editing
functions. For example, the SeaMonkey browser from Mozilla
Foundation includes an HTML editor;
QuickTime Player can be
used to cut and paste parts of video; iPhoto supports a number
of photo editing operations. Conversely, in most cases media
development (or content creation) software such as Word
or PowerPoint is used to both develop and access content. (This
co-existence of authoring and access functions is an important
distinguishing feature of software culture.) If we visit the websites
of popular makers of these software applications such as Adobe
and Autodesk, we will nd that these companies may break their
products by market (web, broadcast, architecture, and so on) or
use sub-categories such as consumer and pro. This is as good
as it getsanother reason why we should focus our theoretical
tools on interrogating media software.
While I will focus on media applications for creating and
accessing content (i.e. media artifacts), cultural software
also includes tools and services that are specically designed
for communication and sharing of information and knowledge,
i.e. social software (categories 34 in my list). The examples
include search engines, web browsers, blog editors, email applica-
tions, instant messaging applications, wikis, social bookmarking,
social networks, virtual worlds, and prediction markets. The
familiar names include Facebook, the family of Google products
(Google Web search, Gmail, Google Maps, Google+, etc.), Skype,
MediaWiki, and Blogger. However, since at the end of the 2000s,
numerous software apps and services started to include email, post,
9781623568177_txt_print.indd 27 29/04/2013 11:50
and chat functions (often via a dedicated Share menu), to an
extent, all software became social software.
Of course, people do not share everything online with others
at least, not yet and not everybody. Therefore, we should also
include software tools for personal information management such
as project managers, database applications, and simple text editors
or note-taking apps that are included with every computer device
being sold.
These and all other categories of software shift over time. For
instance, during the 2000s the boundary between personal infor-
mation and public information has been recongured as people
started to routinely place their media on media sharing sites, and
also communicate with others on social networks.
In fact, the whole reason behind the existence of social media
and social networking services and hosting websites is to erase this
boundary as much as possible. By encouraging users to conduct
larger parts of their social and cultural lives on their sites, these
services can both sell more ads to more people and ensure the
continuous growth of their user base. With more of your friends
using a particular service and ofering more information, media,
and discussions there, you are more likely to also join that service.
As many of these services began to ofer more and more advanced
media editing and information management tools along with their
original media hosting and communication and social networking
functions, they did manage to largely erase another set of boundaries
(from the PC era): those between application programs, operating
system, and data. Facebook in particular was very aggressive in
positioning itself as a complete social platform which can replace
various stand-alone communication programs and services.
Until the rise of social media and the proliferation of mobile
media platforms, it was possible to study media production,
dissemination, and consumption as separate processes. Similarly,
we could usually separate production tools, distribution technol-
ogies, and media access devices and platformsfor example, the
TV studio, cameras, lighting, and editing machines (production),
transmission systems (distribution), and television sets (access).
Social media and cloud computing in general erase these bound-
aries in many cases and at the same time introduce new ones
(client/server, open access/commercial). The challenge of software
studies is to be able to use terms such as content and software
9781623568177_txt_print.indd 28 29/04/2013 11:50
application while always keeping in mind that the current social
media/cloud computing paradigms are systematically reconguring
the meaning of these terms.
Since creation of interactive media often involves writing some
original computer code, the programming environments also can be
considered under cultural software. Moreover, the media interfaces
themselvesicons, folders, sounds, animations, vibrating surfaces,
and touch screensare also cultural software, since these interfaces
mediate peoples interactions with media and other people. I will
stop here but this list can easily be extended to include additional
categories of software as well.
The interface category is particularly important for this book.
I am interested in how software appears to usersi.e. what
functions it ofers to create, share, reuse, mix, create, manage, share
and communicate content, the interfaces used to present these
functions, and assumptions and models about a user, her/his needs,
and society encoded in these functions and their interface design.
These functions ofered by an application are embedded in
application commands and tools. They dene what you can do
with a given app, and how you can do it. This is clear; but I
need to make one important point about interfaces to avoid any
confusion. Many people still think that contemporary computer
devices use a Graphical User Interface (GUI). In reality, the original
GUI of the early 1980s (icons, folders, menus) has been gradually
extended to include other media and senses (sounds, animations,
and vibration feedback which may accompany user interactions on
a mobile device, voice input, multi-touch gesture interfaces, etc.)
This is why the term media interface (used in the industry) is a
more accurate description of how interfaces work today. The term
accurately describes interfaces of computer operating systems such
as Windows and Mac OS, and mobile OS such Android and iOS; it
is even applicable to interfaces of game consoles and mobile phones,
as well as interactive stores
or museum installations which use all
types of media besides graphics to communicate with the users.
For examples, see Nanikas projects for Nokia and Diesel, http://www.nanikawa.
com/; Audi City in London, opened 2012.
For example, see interactive installations at the Nobel Peace Center in Oslo:
Nobel Chamber, Nobel Field, and Nobel Electronic Wall Papers, http://www.
9781623568177_txt_print.indd 29 29/04/2013 11:50
I also need to comment on the media/content vs. data/infor-
mation/knowledge categories used to organize my list of types of
cultural software above. As with many other categories that I will
use in this book, I think of them as marking the two parts of the
same continuous dimension rather than as being discrete either/or
boxes. A feature lm is a good example of the rst category, and
an Excel spreadsheet represents the second categorybut between
such clear-cut examples, there are numerous other cases which
are both. For example, if I make an information visualization of
the data in the spreadsheet, this visualization now ts equally into
both categories. It is still data, but data represented in a new
way which allows us to arrive at insights and knowledge. It also
becomes a piece of visual media which appeals to our senses in the
same way as photographs and paintings do.
The reason that our society places these two sets of terms in
opposition has to do with the histories of the media and infor-
mation industries. Modern media is the result of the technologies
and institutions which developed between the second half of the
eighteenth and rst half of the twentieth centuries: large-scale
newspaper, magazine and book publishing, photography, cinema,
radio, television and the record industry. Data comes from a
number of separate professional elds with distinct histories: social
statistics, economics, business management, and nancial markets.
It is only in the beginning of the twenty-rst century that data leaves
professional domains to become of interest to society at large. Data
becomes sexy and hip, with governments and cities creating
their own data portals (for example, data.gov and data.gov.uk),
visualizations of data entering exhibitions of major museums such
as MOMA (Design and Elastic Mind, 2008), the computer nerds
becoming heroes of Hollywood lms (Social Network, 2010),
and Google Analytics, Facebook, YouTube and Flickr all ofering
detailed data about your website or media sharing account. Of
course, since media software operations (as well as any other
computer processing of media for research, commercial or artistic
purposes) are only possible because the computer represents media
as data (discrete elements such as pixels, or equations dening
vector graphics in vector les such as EPS), the development of
media software and its adoption as the key media technology
(discussed in this book) is an important contributor to the gradual
coming together of media and data.
9781623568177_txt_print.indd 30 29/04/2013 11:50
Software includes many other technologies and types, and
computers and computer devices also perform lots of other functions
besides creating and playing media. And of course, software needs
hardware to run; and networks are also an essential part of our
digital culture. Therefore, my focus on software applications for
creating, editing, and playing media is likely to annoy some people.
Not everybody uses Photoshop, Flash, Maya, and other applications
to create media. A signicant number of people work with media
by writing their own computer programs and scripts, or modifying
programs written by others. These are programmers responsible
for the coding of websites, web applications, and other interactive
applications, software artists, computer scientists working on the
development of new algorithms, students using Processing and
other high-level media programming languages, and other groups.
All of them may ask me why I single out software in the form of
consumer products (i.e. applications)as opposed to the activity of
programming? And what about the gradual democratization of
software development and the gradual increase in the number
of culture professionals and students who can program or write
scripts? Should I not put my energy into promoting programming
rather than explaining applications?
The reason for my choice is my commitment to understand the
mainstream cultural practices rather than to emphasize (as many
cultural critics do) the exceptions, no matter how progressive they
may be. Although we do not have an exact number, I assume
that the number of people who work in media and who can also
program is tiny in comparison to the army of application users.
Today, a typical professional graphic designer, lm editor, product
designer, architect, music artistand certainly a typical person
uploading videos to YouTube or adding photos and video on her/
his blogcan neither write nor read software code. (Being able to
read and modify HTML markup, or copy already pre-packaged
lines of Javascript code is very diferent from programming.)
Therefore, if we want to understand how software has already
re-shaped media both conceptually and practically, we have to
take a close look at the everyday tools used by the great majority
of both professional and non-professional usersi.e. application
software, web-based software, and, of course, mobile apps. (This
book highlights the rst category at the expense of the second and
the thirdbecause at this point, creation of professional media still
9781623568177_txt_print.indd 31 29/04/2013 11:50
requires applications running on a laptop or desktop, often with a
signicant amount of RAM and large hard drives; and also because
currently web-based and mobile software are still evolving quite
rapidly in contrast to desktop applications such as Photoshop and
Final Cut which change only incrementally from release to release).
Any denition is likely to delight some people and to annoy
others. Therefore, I also would like to address another likely
objection to the way I dened the term cultural software (with
media software being its subset). The term culture is not
reducible to separate media and design objects which may exist
as les on a computer and/or as executable software programs or
scripts. It includes symbols, meanings, values, language, habits,
beliefs, ideologies, rituals, religion, dress and behavioral codes,
and many other material and immaterial elements and dimensions.
Consequently, cultural anthropologists, linguists, sociologists, and
many humanists may be annoyed at what may appear as an
uncritical reduction of all these dimensions to a set of tools for
creating and playing media les.
Am I saying that today culture is equated with a particular
subset of application software and the media objects and experi-
ences that can be created with their help? Of course not. However,
what I am sayingand what I hope this book explicates in more
detailis that at the end of the twentieth century humans have
added a fundamentally new dimension to everything that counts as
culture. This dimension is software in general, and application
software for creating and accessing content in particular.
I am using the metaphor of a new dimension on purpose. That
is, cultural software is not simply a new objectno matter
how large and importantwhich has been dropped into the
space which we call culture. Thus, it would be imprecise to
think of software as simply another term which we can add to
the set which includes music, visual design, built spaces, dress
codes, languages, food, club cultures, corporate norms, ways of
talking and using a body, and so on. And while we can certainly
study the culture of softwareprogramming practices, values
and ideologies of programmers and software companies, the
cultures of Silicon Valley and Bangalore, etc.if we only do this,
we will miss the real importance of software. Like the alphabet,
mathematics, printing press, combustion engine, electricity, and
integrated circuits, software re-adjusts and re-shapes everything it
9781623568177_txt_print.indd 32 29/04/2013 11:50
is applied toor at least, it has a potential to do this. Just as adding
a new dimension adds a new coordinate to every point in space,
adding software to culture changes the identity of everything
that a culture is made from. (In this respect, software is a perfect
example of what McLuhan meant when he wrote, the message of
any medium or technology is the change of scale or pace or pattern
that it introduces into human afairs.
To summarize: our contemporary society can be characterized
as a software society and our culture can be justiably called a
software culturebecause today software plays a central role in
shaping both the material elements and many of the immaterial
structures that together make up culture.
From documents to performances
The use of software re-congures most basic social and cultural
practices and makes us rethink the concepts and theories we
developed to describe them. As one example of this, consider the
modern atom of cultural creation, transmission, and memory:
a document, i.e. some content stored in a physical form,
that is delivered to consumers via physical copies (books, lms,
audio record), or electronic transmission (television). In software
culture, we no longer have documents, works, messages
or recordings in twentieth-century terms. Instead of xed
documents that could be analyzed by examining their structure and
content (a typical move of the twentieth-century cultural analysis
and theory, from Russian Formalism to Literary Darwinism),
we now interact with dynamic software performances. I use
the word performance because what we are experiencing is
constructed by software in real time. So whether we are exploring
a dynamic website, playing a video game, or using an app on a
mobile phone to locate particular places or friends nearby, we
are engaging not with pre-dened static documents but with the
dynamic outputs of a real-time computation happening on our
Marshall McLuhan, Understanding Media: The Extensions of Man (New York:
McGraw Hill, 1964), quoted in New Media Reader, Noah Wardrip-Fruin and Nick
Montfort (eds) (The MIT Press, 2003), p. 203.
9781623568177_txt_print.indd 33 29/04/2013 11:50
device and/or the server. Computer programs can use a variety
of components to create these performances: design templates,
les stored on a local machine, media from the databases on the
network server, the real-time input from a mouse, touch screen,
joystick, our moving bodies, or some other interface. Therefore,
although some static documents may be involved, the nal media
experience constructed by software usually does not correspond to
any single static document stored in some media. In other words, in
contrast to paintings, literary works, music scores, lms, industrial
designs, or buildings, a critic cannot simply consult a single le
containing all of the works content.
Even in such seemingly simple cases as viewing a PDF document
or opening a photo in a media player, we are already dealing with
software performancessince it is software which denes the
options for navigating, editing, and sharing the document, rather
than the document itself. Therefore examining the PDF le or a
JPEG le the way twentieth-century critics would examine a novel,
a movie, or a TV program will only tell us some things about
the experience we get when we interact with this document via
softwarebut not everything. This experience is equally shaped by
the interface and the tools provided by software. This is why the
examination of the tools, interfaces, assumptions, concepts, and the
history of cultural softwareincluding the theories of its inventors
who in the 1960s and 1970s have dened most of these concepts
is essential if we are to make sense of contemporary media.
This shift in the nature of what constitutes a media document
also calls into question well-established cultural theories that
depend on this concept. Consider the intellectual paradigm that
dominated the study of media since the 1950sthe trans-
mission view of culture developed in Communication Studies.
Communication scholars have taken the model of information
transmission formulated by Claude Shannon in his 1948 article A
Mathematical Theory of Communication (1948)
and his subse-
quent book published with Warren Weaver in 1949,
and applied
C. E. Shannon, A Mathematical Theory of Communication, Bell System
Technical Journal, vol. 27, pp. 379423, 62356, July, October, 1948, http://
Claude E. Shannon and Warren Weaver, The Mathematical Theory of
Communication (University of Illinois Press, 1949).
9781623568177_txt_print.indd 34 29/04/2013 11:50
its basic model of communication to mass media. The paradigm
described mass communication (and sometimes culture in general)
as a communication process between the authors who create and
send messages and the audiences that receive them. According to
this paradigm, the messages were not always fully decoded by the
audiences for technical reasons (noise in transmission) or semantic
reasons (they misunderstood the intended meanings).
Classical communication theory and media industries considered
such partial reception a problem; in contrast, in his inuential 1980
article Encoding/decoding
the founder of British Cultural
Studies, Stuart Hall, argued that the same phenomenon is positive.
Hall proposed that the audiences construct their own meanings
from the information they receive. Rather than being a commu-
nication failure, the new meanings are active acts of intentional
reinterpretation of the sent messages. But both the classical commu-
nication studies and cultural studies implicitly took for granted that
the message was something complete and deniteregardless of
whether it was stored in physical media (e.g. magnetic tape) or
created in real time by a sender (a live TV broadcast). Thus, the
receiver of communication was assumed to read all of the adver-
tising copy, see a whole movie, or listen to the whole song and only
after that s/he would interpret it, misinterpret it, assign his/her own
meanings, appropriate it, remix it, etc.
While this assumption has already been challenged by the intro-
duction of the DVR (digital video recorder) in 1999, which led
to the phenomenon of time shifting, it simply does not apply to
interactive software-driven media. The interfaces of media access
applications, such as web browsers and search engines, the hyper-
linked architecture of the World Wide Web, and the interfaces
of particular online media services ofering massive numbers of
media artifacts for playback preview and/or purchase (Amazon,
Google Play, iTunes, Rhapsody, Netix, etc.), encourage people
to browse, quickly moving both horizontally between media
(from one search result to the next, from one song to another, etc.)
and vertically, through the media artifacts (e.g., from the contents
listing of a music album to a particular track). They also made it
easy to start playing/viewing media at an arbitrary point, and to
Stuart Hall, Encoding/decoding, in Culture, Media, Language, ed., Centre for
Contemporary Cultural Studies (London: Hutchinson, 1980).
9781623568177_txt_print.indd 35 29/04/2013 11:50
leave it at any point. In other words, the message that the user
receives is not just actively constructed by him/her (through a
cognitive interpretation) but also actively managed (dening what
information s/he is receiving and how).
It is at least as important that when a user interacts with a
software application that presents media content, this content often
does not have any xed nite boundaries. For instance, a user of
Google Earth is likely to experience a diferent earth every time
s/he is accessing the application. Google could have updated some
of the satellite photographs or added new Street Views; new 3D
buildings, new layers, and new information on already existing
layers were also likely to be added. Moreover, at any time a user of
the application can load more geospatial data created by other users
and companies by either selecting one of the options in the Add
menu (Google Earth 6.2.1 interface), or directly opening a KLM le.
Google Earth is a typical example of a new type of media enabled
by the weban interactive document which does not have all of its
content pre-dened. Its content changes and grows over time.
In some cases this may not afect in any signicant way the
larger messages communicated by the software application,
web service, game, or other type of interactive media. For example,
Google Earths built-in cartographic convention of representing the
Earth using the General Perspective Projection (a particular map
projection method of cartography) does not change when users add
new content and turn on and of map layers. The message of this
representation is always present.
However, since a user of Google Earth can also add his/her own
media and information to the base representation provided by the
application, creating complex and media rich projects on top of
existing geoinformation, Google Earth is not just a message. It
is a platform for users to build on. And while we can nd some
continuity here with the users creative reworking of commercial
media in the twentieth centurypop art and appropriation, music
remixes, slash ction and video,
and so on, the diferences are
larger than the similarities.
http://en.wikipedia.org/wiki/Google_earth#Technical_specifications (March 14,
See, for instance, Constance Penley, Feminism, Psychoanalysis, and the Study of
Popular Culture, in Cultural Studies, ed. Lawrence Grossberg (Routledge, 1992).
9781623568177_txt_print.indd 36 29/04/2013 11:50
This shift from messages to platforms was in the center of the
webs transformation around 20046. The result was named Web
2.0. The 1990s websites presenting particular content created by
others (and thus, communicating messages) were supplemented
by social networks and social media sites where the users can
share, comment on, and tag their own media. The Wikipedia
article on Web 2.0 describes these diferences as follows: A Web
2.0 site allows users to interact and collaborate with each other in
a social media dialogue as creators (prosumers) of user-generated
content in a virtual community, in contrast to websites where
users (consumers) are limited to the passive viewing of content
that was created for them. Examples of Web 2.0 include social
networking sites, blogs, wikis, video sharing sites, hosted services,
web applications, mashups and folksonomies.
For example, to
continue with the Google Earth example, users added many types
of global awareness information, including fair trade certication,
Greenpeace data, and United Nations Millennium Development
Goals Monitor.
In another example, you can incorporate Google
Maps, Wikipedia, or content provided by most other large web
2.0 sites directly in your web mashupan even more direct way of
taking the content provided by web services and using it to craft
your own custom platforms.
The wide adoption of Web 2.0 services along with various
web-based communication tools (online discussion forums
about all popular software, collaborative editing on Wikipedia,
Twitter, etc.) enables quick identications of omissions, selec-
tions, censorship and other types of bad behavior by software
publishersanother feature which separates content distributed
by web-based companies from mass media of the twentieth
century. For example, every article on Wikipedia about a Web 2.0
service includes a special section about controversies, criticism, or
In many cases, people can also use alternative open source
equivalents of paid and locked applications. Open source and/
or free software (not all free software is open source) often
allow for additional ways of creating, remixing and sharing
both content and new software additions. (This does not mean
http://en.wikipedia.org/wiki/Web_2.0 (March 14, 2012).
http://en.wikipedia.org/wiki/Google_earth (March 14, 2012).
9781623568177_txt_print.indd 37 29/04/2013 11:50
that open source software always uses diferent assumptions and
technologies than the commercial software.) For example, one
can choose to use a number of alternatives to Google Maps and
Google EarthOpenStreetMap, Geocommons, WorldMap, and
others which all have open source or free software licenses.

(Interestingly, commercial companies also often use data from
such free collaboratively created systems because they contain
more information than the companies own systems. OpenStreet
Map, which by early 2011 had 340,000 contributors,
is used
by Flickr and Foursquare.
) A user can also examine the code of
open-source software to fully understand its assumptions and key
Continuously changing and growing content of web services
and sites; variety of mechanism for navigation and interaction; the
abilities to add ones own content and mashup content from various
sources together; architectures for collaborative authoring and
editing; mechanisms for monitoring the providersall these mecha-
nisms clearly separate interactive networked software-driven media
from twentieth-century media documents. But even when a user is
working with a single local media document that is stored in a single
computer le (a rather rare situation these days), such a document
mediated through software interface has a diferent identity from a
twentieth-century media document. The users experience is only
partly dened by the les content and its organization. The user
is free to navigate the document, choosing both what information
to see and the sequence in which s/he is seeing it. And while old
media (with the exception of twentieth-century broadcasting)
also provided this random access, the interfaces of software-driven
media players/viewers provide many additional ways for browsing
media and selecting what and how to access.
For example, Adobe Acrobat can display thumbnails of every
page in a PDF document; Google Earth can quickly zoom in and out
from the current view; online digital libraries, databases and repos-
itories containing scientic articles and abstracts such as the ACM
http://geocommons.com, http://www.openstreetmap.org, http://worldmap.
http://en.wikipedia.org/wiki/Counter-mapping#OpenStreetMap (March 27, 2012).
Data (March 27, 2012).
9781623568177_txt_print.indd 38 29/04/2013 11:50
Digital Library, IEEE Xplore, PubMed, Science Direct, SciVerse
Scopus, and Web of Science show articles which contain references
to the one you currently selected. Most importantly, these new
tools and interfaces are not hard-wired to the media documents
themselves (such as a random access capacity of a printed book)
or media access machines (such as a radio); instead they are part
of the separate software layer. This media architecture enables easy
addition of new navigation and management tools without any
change to the documents themselves. For instance, with a single
click, I can add sharing buttons to my blog, thus enabling new
ways of circulation for its content. When I open a text document
in Mac OS X Preview media viewer, I can highlight, add comments
and links, draw, and add thought bubbles. Photoshop allows me to
save my edits on separate adjustment layers, without modifying
the original image. And so on.
Why the history of cultural software does
not exist
Bcxxoe onncanne mnpa cnntno orcraer or ero pasnnrnx.
(Translation from Russian: Every description of the world
substantially lags behind its actual development.)
Tax Karmma, VJ on MTV.ru, 2008.
We live in a software culturethat is, a culture where the
production, distribution, and reception of most content is mediated
by software. And yet, most creative professionals do not know
anything about the intellectual history of software they use daily
be it Photoshop, Illustrator, GIMP, Final Cut, After Efects, Blender,
Flame, Maya, MAX, or Dreamweaver.
Where does contemporary cultural software came from? How
were its metaphors and techniques arrived at? And why was it
developed in the rst place? Currently most prominent computer
and web companies have been extensively covered in media, so
their history is relatively well-known (for instance, Facebook,
Google, and Apple). But this is only the tip of the iceberg.
The history of media authoring and editing software remains
pretty much unknown. Despite the common statements that the
9781623568177_txt_print.indd 39 29/04/2013 11:50
digital revolution is at least as important as the invention of the
printing press, we are largely ignorant of how the key part of this
revolutioni.e., media softwarewas invented. When you think
about this, it is unbelievable. People in the business of culture know
about Gutenberg (printing press), Brunelleschi (perspective), The
Lumire Brothers, Grifth and Eisenstein (cinema), Le Corbusier
(modern architecture), Isadora Duncan (modern dance), and Saul
Bass (motion graphics). (If you happen not to know one of these
names, I am sure that you have other cultural friends who do).
And yet, even today, relatively few people have heard of J. C. R.
Licklider, Ivan Sutherland, Ted Nelson, Douglas Engelbart, Alan
Kay, and their collaborators who, between approximately 1960
and 1978, gradually turned the computer into the cultural machine
it is today.
Remarkably, the history of cultural software as a discrete
category does not yet exist. What we have are a number of largely
biographical books about some of the key individual gures, and
research labs such as Xerox PARC or MIT Media Labbut no
comprehensive synthesis which would trace the genealogical tree
of media tools.

And we also do not have any detailed studies which
would relate the history of cultural software to the history of
media, media theory, or history of visual culture.
Modern art institutionsmuseums such as the MOMA and
the Tate, art book publishers such as Phaidon and Rizzoli, etc.
promote the history of modern art. Hollywood is similarly proud
of its own historythe stars, the directors, the cinematographers,
and the classical lms. So how can we understand the neglect of
the history of cultural computing by our cultural institutions and
computer industry itself? Why, for instance, does Silicon Valley
not have a museum for cultural software? (The Computer History
museum in Mountain View, California has an extensive permanent
exhibition which is focused on hardware, operating systems, and
programming languagesbut not on the history of software.
I believe that the major reason has to do with economics.
Originally misunderstood and ridiculed, modern art has eventually
become a legitimate investment categoryin fact, by middle of
the 2000s, the paintings of a number of twentieth-century artists
http://www.mtv.ru/air/vjs/taya/main.wbp (February 21, 2008).
9781623568177_txt_print.indd 40 29/04/2013 11:50
were selling for more money than the works of the most famous
classical artists. Similarly, Hollywood continues to receive prots
from old movies as it reissues them in new formats (VHS, DVD,
HD, Blu-ray disks, etc). What about the IT industry? It does not
derive any prots from the old softwareand therefore it does
nothing to promote its history. Of course, contemporary versions
of Microsoft Word, Adobe Photoshop, Autodesk AutoCAD, and
many other popular cultural applications were built on the rst
versions, which often date from the 1980s, and the companies
continue to benet from the patents they led for new technologies
used in these original versionsbut, in contrast to the video games
from the 1980s, these early software versions are not treated as
separate products which can be re-issued today. (In principle, I
can imagine the software industry creating a whole new market
for old software versions or applications which at some point
were quite important but no longer exist todayfor instance,
Aldus Pagemaker. In fact, given that consumer culture systemati-
cally exploits adults nostalgia for the cultural experiences of their
teenage and youth years, it is actually surprising that early software
versions were not seen as a market opportunity. If I used MacWrite
and MacPaint daily in the middle of the 1980s, or Photoshop 1.0
and 2.0 in 19903, I think these experiences would be as much
part of my cultural genealogy as the movies and art I saw at that
time. Although I am not necessarily advocating the creation of yet
another category of commercial products, if early software was
widely available in simulation, it would catalyze cultural interest
in software similar to the way in which wide availability of early
computer games, recreated for contemporary mobile platforms,
fuels the eld of video game studies.
Since most theorists so far have not considered cultural software
as a subject of its own, distinct from social media, social
networks, new media, media art, the internet, interac-
tivity, and cyberculture, we lack not only a conceptual history
of media editing software but also systematic investigations of the
roles of software in media production. For instance, how did the
adoption of the popular animation and compositing application
After Efects in the 1990s reshape the language of moving images?
How did the adoption of Alias, Maya and other 3D packages by
architectural students and young architects in the same decade
similarly inuence the language of architecture? What about the
9781623568177_txt_print.indd 41 29/04/2013 11:50
co-evolution of Web design tools and the aesthetics of websites
from the bare-bones HTML in 1994 to visually rich Flash-driven
sites ve years later, and responsive web design in the early 2010s?
You will nd frequent mentions and short discussions of these
and similar questions in articles and conference talks, but as far
as I know, there has been no book-length study about any of these
subjects. Often, books on architecture, motion graphics, graphic
design and other design elds will briey discuss the importance
of software tools in facilitating new possibilities and opportunities,
but these discussions are not usually further developed.
In summary, a systematic examination of the connections
between the workings of contemporary media software and the
new communication languages in design and media (including
graphic design, web design, product design, motion graphics,
animation, and cinema) has not yet been undertaken. Although
this book alone cannot do it all, I hope that it will provide some
general models of how such connections can be teased outas well
as provide a detailed analysis of how software use redened certain
cultural areas (e.g., motion graphics and visual design).
By focusing on the theory of software for media design, this
book aims to complement the work of a few other theorists that
have already examined software responsible for game platforms
and design (Ian Bogost, Nick Montfort), and electronic literature
(Noah Wardrip-Fruin, Matthew Kirschenbaum).
In this respect, the related elds of code studies and platform
studies being developed by Mark Marino,
Nick Montfort, Ian
Bogost and others are playing a very important role. According
to Marino (and I completely agree), the three elds of software
studies, code studies, and game studies complement each other:
Critical code studies is an emerging eld related to software
studies and platform studies, but its more closely attuned to the
code itself of a program rather than the programs interface and
usability (as in software studies) or its underlying hardware (as in
platform studies).

http://chnm2011.thatcamp.org/05/24/session-proposal-critical-code studies/ (July
14, 2011).
9781623568177_txt_print.indd 42 29/04/2013 11:50
Summary of the books narrative
Between the early 1990s and the middle of the 2000s, media
software has replaced most of the other media technologies that
emerged in the nineteenth and twentieth centuries. Most contem-
porary media is created and accessed via cultural softwareand
yet, surprisingly, few people know about its history. What was
the thinking and motivation of people who between 1960 and
the late 1970s created the concepts and practical techniques that
underlie todays cultural software? How does the shift to software-
based production methods in the 1990s change our concepts of
media? How have interfaces and the tools of content devel-
opment software reshaped and continued to shape the aesthetics
and visual languages we see in contemporary design and media?
These are the key questions that I take up in this book.
My aim is not provide a comprehensive history of cultural
software in general, or media authoring software in particular. Nor
do I aim to discuss all the new creative techniques media software
enables across dozens of cultural elds. Instead, I will trace a
particular path through this history that will take us from 1960 to
today and which will pass through some of its most crucial points.
In the following I summarize this narrative and also introduce
some of the key concepts developed in each part of the book.
Part 1 looks at the 1960s and 1970s. While new media theorists
have spent considerable eforts in trying to understand the relation-
ships between digital media and older physical and electronic
media, the important sourcesthe writing and projects by Ivan
Sutherland, Douglas Engelbart, Ted Nelson, Alan Kay, and other
pioneers of cultural software working in these decadesstill remain
largely unexamined. What were their reasons for inventing the
concepts and techniques that today make it possible for computers
to represent, or remediate other media? Why did these people
and their colleagues work to systematically turn a computer into
a machine for media creation and manipulation? These are the
questions that I take in Part 1, which explores them by focusing
on the ideas and work of the key protagonist of cultural software
movementAlan Kay. (It is certainly possible to construct a more
exclusive or an alternative history which will pay equal attention
to dozens of brilliant people who worked with these people and
9781623568177_txt_print.indd 43 29/04/2013 11:50
who, together, invented all the details which form the DNA of
contemporary media softwarefor instance, Bob Taylor, Charles
Thacker, John Warnock, and others working at Xerox PARC in
the 1970s; or the people who contributed to the design of the rst
However, since we do not yet even have a theoretical
analysis of how the ideas of the most well-known gures of the
1960s collectively changed media, this book will start with these
gures, and the analysis of their theoretical writings.)
I suggest that Kay and cultural software pioneers aimed to create
a particular kind of new mediarather than merely simulating the
appearances of old ones. These new media use already existing
representational formats as their building blocks, while adding
many previously nonexistent properties. At the same time, as
envisioned by Kay, these media are expandablethat is, users
themselves should be able to easily add new properties, as well as
to invent new media. Accordingly, Kay calls computers the rst
metamedium whose content is a wide range of already-existing
and not-yet-invented media.
The foundations necessary for the existence of such metamedium
were established between the 1960s and the late 1970s. During this
period, most previously available physical and electronic media
were systematically simulated in software, and a number of new
media were also invented. This development takes us from the very
interactive design programIvan Sutherlands Sketchpad (1962)
to the commercial desktop applications that made software-based
media authoring and design widely available to members of
diferent creative professions and, eventually, media consumers as
wellAutoCAD (1982), Word (1984), PageMaker (1985), Alias
(1985), Illustrator (1987), Director (1987), Photoshop (1989),
After Efects (1993), and others. (These PC applications were paral-
leled by much more expensive systems for professional markets
such as the TV and video industries which got Paintbox in 1981,
Harry in 1985, Avid in 1989, and Flame in 1992.)
So what happens next? Did Kays theoretical formulations as
articulated in 1977 accurately predict the developments of the
next thirty years, or have there been new developments that his
concept of metamedium did not account for? Today we do
For the stories that document the inventions by dozens of people of the multiple
technologies that made up the original Macintosh, see www.folklore.com
9781623568177_txt_print.indd 44 29/04/2013 11:50
indeed use a variety of previously existing media simulated in
software as well as new previously non-existent media types. Both
have been continuously extended with new properties. Do these
processes of invention and amplication take place at random,
or do they follow particular paths? In other words, what are the
key mechanisms responsible for the extension of the computer
Parts 2 and 3 are devoted to these questions. They look at the
number of diferent mechanisms which drove development and
expansion of the computer metamedium, with the focus on the
1990s when media software was gradually adopted in all areas of
professional media production. I use three diferent concepts to
describe these developments and the new aesthetics of visual media
which developed in the second part of the 1990s after the processes
of adoption reached sufcient speed. These three concepts are
media hybridization, evolution, and deep remix. Part 2 develops
the theoretical analysis of this second stage of metamedium devel-
opment, illustrating it with a number of examples drawn from
diferent genres of digital media. Part 3 focuses in detail on the
use of software for visual design (motion graphics and graphics
design), analyzing the relationships between the new aesthetics of
moving and still images and compositions, and the operations and
interfaces of software used to create them such as After Efects.
I argue that in the process of the translation from physical and
electronic media technologies to software, all individual techniques
and tools that were previously unique to diferent media met
within the same software environment. This meeting had funda-
mental consequences for human cultural development and for the
media evolution. It disrupted and transformed the whole landscape
of media technologies, the creative professions that use them, and
the very concept of media itself.
Once they were simulated in a computer, previously incom-
patible techniques of diferent media begin to be combined in
endless new ways, leading to new media hybrids, or, to use a
biological metaphor, new media species. As just one example
among countless others, think, for instance, of the popular Google
Earth application, combining techniques of traditional mapping,
the concepts from the eld of Geographical Information Systems
(GIS), 3D computer graphics and animation, social software,
search, and other elements and functions. In my view, this ability to
9781623568177_txt_print.indd 45 29/04/2013 11:50
combine previously separate media techniques represents a funda-
mentally new stage in the history of human media, human semiosis,
and human communication, enabled by its softwarization.
I describe this new stage in media evolution using the concept
of hybridity. In the rst stage, most existing media were simulated
in a computer and a number of new types of media that can only
be realized in a computer were invented. In the second stage, these
simulated and new mediums started exchanging properties and
To distinguish these processes from more familiar remixes, I
introduce the new term deep remixability. Normally a remix is
a combination of content from a single medium (like in music
remixes), or from a few mediums (like Anime Music Video works
which combine content from anime and music video). However,
the software production environment allows designers to remix
not only the content of diferent media types, but also their funda-
mental techniques, working methods, and ways of representation
and expression.
While today hybridization and deep remix can be found at work
in all areas of culture where software is used, I focus on particular
area to demonstrate how it functions in detail. This area is visual
design in general, and motion graphics in particular. Motion graphics
is a dynamic part of contemporary culture, which, as far as I know,
has not yet been theoretically analyzed in detail anywhere. Although
selected precedents for contemporary motion graphics can already be
found in the 1950s and 1960s in the works by Saul Bass and Pablo
Ferro, its exponential growth from the middle of the 1990s is directly
related to the adoption of software for moving image design
specically, After Efects software released by Adobe in 1993. Deep
remixability is central to the aesthetics of motion graphics. That is,
the larger proportion of motion graphics projects done today around
the world derive their aesthetic efects from combining diferent
techniques and media traditionsanimation, drawing, typography
photography, 3D graphics, video, etc.in new ways. As a part of
my analysis, I look at how the typical software-based production
workow in a contemporary design studiothe ways in which a
project moves from one software application to anothershapes the
aesthetics of motion graphics, and visual design in general.
The next major wave of computerization of culture has to do
with diferent types of softwaresocial networks, social media
9781623568177_txt_print.indd 46 29/04/2013 11:50
services, and apps for mobile platforms. The wave of social
networks and social media started slowly, erupted in 20052006
(Flickr, YouTube) and continues to move forward and expand
its reach. The 1990s media revolution impacted professional
creatives; the 2000s media revolution afected the rest of usi.e.
the hundreds of millions who use Facebook, Twitter, Firefox,
Safari, Google Search and Maps, Flickr, Picasa, Vimeo, Blogger,
and numerous apps and services available on mobile platforms.
Because we are still in the middle of social media difusion, with
some popular social media services going out of favor and others
gaining speed (for example, think of the fate of MySpace), and
the social functionality of software still expanding, I decided
that ofering the detailed theoretical analysis of this new wave
would be premature. (This became clear after I started editing the
part about social media which I originally had in the rst book
draft, and realized that some of the social media services I was
analyzing in detail no longer exist .) Instead, I am focusing on
tracing the fundamental developments which made possible and
shaped digital media before its social explosion: the ideas about
the computer as a machine for media generation and editing of
the 1960s1970s, their implementation in the media applica-
tions in the 1980s1990s, and the transformation of visual media
languages which quickly followed.
To be more precise, we can frame this history between 1961 and
1999. In 1961, Ivan Sutherland at MIT designed Sketchpad, which
became the rst computer design system shown to the public. In
1999, After Efects 4.0 introduced Premiere import,
5.5 added vector shapes,
and Apple showed the rst version of
Final Cut Pro
in short, the current paradigm of interoperable
media authoring and editing tools capable of creating professional
media without special hardware beyond the of-the-shelf computer
was nalized. And while professional media tools continued to
evolve after this period, the changes so far have been incremental.
Similarly, the languages of professional visual media created with
this software did not change signicantly after their radical trans-
formation in the second part of the 1990s.
http://en.wikipedia.org/wiki/After_Effects#History (July 7, 2011).
http://en.wikipedia.org/wiki/Adobe_Photoshop_release_history (July 7, 2011).
http://en.wikipedia.org/wiki/Final_Cut_Pro#History (July 7, 2011).
9781623568177_txt_print.indd 47 29/04/2013 11:50
To illustrate this continuity, the examples of particular media
projects that I will analyze will be drawn from both the 1990s and
the 2000s. However, when discussing interfaces and commands
of media applications, I will use the recent versions in order to
make the discussion as relevant as possible for the software users.
Accordingly, I will also take into account the addition of social
media capacities in all consumer-level media software that took
place at the end of the 2000s (e.g., the share menu in iPhoto).
And because the hybridization mechanism is not limited to profes-
sional media software and professionally created media, but also
plays the key role in evolution of social web software and services,
I will include prominent examples of such services, for example,
Google Earth.
I also need to comment on my choices of particular media
software applications as examples. I have chosen to focus on the
desktop applications for media authoring most widely used today
Photoshop, Illustrator, InDesign, Dreamweaver, After Efects, Final
Cut, Maya, 3ds Max, Word, PowerPoint, etc. These programs
exemplify diferent categories of media authoring software: image
editing, vector graphics, page layout, web design, motion graphics,
video editing, 3D modeling and animation, word processing, and
presentation. I will also be making references to popular web
browsers (Firefox, Chrome, Internet Explorer), blogging tools
and publishing services (WordPress, Blogger), social networks
(Facebook, Twitter, Google+), media sharing services (Flickr,
Pinterest, YouTube, Vimeo), email services and clients (Gmail,
Microsoft Outlook), web-based ofce suites (Google Docs), and
consumer geographic information systems (Google Earth, Bing
Maps). Since I am interested in how users interact with media,
another key software category for this book is media players
pre-installed on new computers (Windows Media Player, iTunes,
QuickTime) and document viewing applications (Adobe Reader,
Mac OS Preview). As some programs and web services become less
popular and new ones gain market share, the list above may look
somewhat diferent by the time you are reading this, and many
applications may also fully migrate from desktop to the webbut
the categories are likely to remain the same.
Because I want to make my discussions as relevant as possible
to contemporary designers and artists, the names of historically
important programs which are no longer popular or do not exist
9781623568177_txt_print.indd 48 29/04/2013 11:50
will only be mentioned in passing. Examples are QuarkXPress,
WordPerfect, and Macromedia Director. Luckily for us, the two
programs that I will analyze in detailPhotoshop (Chapter 2) and
After Efects (Chapter 5)are as popular today as they were in the
I will not be discussing other types of digital media authoring
and editing systems which were quite important in the 1980s and
1990s. Because during this period graphics capacities of personal
computers were still limited, these systems ran on graphics worksta-
tions (specialized minicomputers) from Silicon Graphics or used
proprietary hardware. Here are examples listed in chronological
order, with the function of the system, company name and year
of the rst release appearing in parentheses: Paintbox (graphics
for broadcast television, Quantel, 1981), Mirage (digital real-time
video efects processor, Quantel, 1982),
Personal Visualizer (3D
modeling and animation, Wavefront, 1988), Henry and Hal (efects
editor and graphics and compositing systems, Quantel, 1992)
Inferno and Flame (compositing for lm and video, Discreet Logic,
In the middle 1990s, Flame together with the SGI workstation
cost $450,000; an Inferno system cost $700,000.
Inferno 5 and
Flame 8 introduced in 2003 had suggested list prices of $571,500
and $266,500, respectively.
Because of these prices, such systems
were only used in television and lm studios or in big video efects
Today the most demanding areas of media production which
involve working with massive amounts of datafeature lms,
feature animations, TV commercialsstill rely on these systems
expensive software. While at the end of the 2000s the companies
started to ofer versions of these programs for PC, Macs, and Linux,
today the highest-end versions still often require special hardware,
and their prices are still quite high. (For example, the 2010 edition
of Autodesk Flame Premium, a suite containing Smoke, Flame and
http://en.wikipedia.org/wiki/Quantel_Mirage (August 23, 2012).
Discreet Logic Inc. History, http://www.fundinguniverse.com/company-
Autodesk, Discreet Delivers inferno 5, ame 8 and int 8, January 23, 2003,
9781623568177_txt_print.indd 49 29/04/2013 11:50
Lustre used for video editing, efects, and color grading was ofered
for $125,000.)
Because I do not expect a typical reader of this book to have
a working experience with these expensive systems, I will not be
referring to them further in this book. However, a more compre-
hensive history of the moving image media of the 1980s1990s
(which I hope somebody will write in the future) will denitely
need to do an archeology and genealogy of these systems and their
Finally, one more explanation is in order. Some readers will
be annoyed that I focus on commercial applications for media
authoring and editing, as opposed to their open source alterna-
tives. For instance, I discuss Photoshop rather than Gimp, and
Illustrator rather than Inkscape. I love and support open source
and free access, and use it for all my work. Starting in 1994, I was
making all my articles available for free download on my website
manovich.net. And when in 2007 I set up a research lab (www.
softwarestudies.com) to start analyzing and visualizing massive
large media datasets, we decided to also follow a free software/
open source strategy, making the tools we develop freely available
and allowing others to modify them.

The reason this book focuses on commercial media authoring
and editing software rather than its open source equivalents is
simple. In almost all areas of software culture, people use free
applications and web services. The examples include web browsers,
web email, social networks, apps available for mobile devices,
and programming and scripting languages. The companies are
not charging for these free applications and services because they
are making money in other ways (advertising, charging for extra
features and services, membership fees, selling devices). However,
in the case of professional tools for media authoring and editing,
commercial software dominates. It is not necessarily better, but
it is simply used by many more people. (For example, entering
Photoshop and Gimp into Google Trends shows that since
2004, the number of searches for the former is about eight times
bigger than for the latter.) Since I am interested in describing the
common user experiences, and the features of media aesthetics
9781623568177_txt_print.indd 50 29/04/2013 11:50
common to millions of works created with the most common
authoring tools that are all commercial products, these are the
products I choose to analyze. And when I analyze tools for media
access and collaboration, I similarly choose the most popular
productswhich in this case includes both free software and
services provided by commercial companies (Safari, Google Earth),
and free open source software (Firefox).
9781623568177_txt_print.indd 51 29/04/2013 11:50
9781623568177_txt_print.indd 52 29/04/2013 11:50
media software
9781623568177_txt_print.indd 53 29/04/2013 11:50
9781623568177_txt_print.indd 54 29/04/2013 11:50
Alan Kays universal
media machine
8.a. A specic kind of artistic technique or means of expression
as determined by the materials used or the creative methods
involved: the medium of lithography.
b. The materials used in a specic artistic technique: oils as a
American Heritage Dictionary, 4th edition
(Houghton Mifin, 2000)
The best way to predict the future is to invent it.
Alan Kay
Appearance versus function
Between its invention in the mid-1940s and the arrival of PCs in
the early 1980s, the digital computer was mostly used for military,
scientic, and business calculations and data processing. It was not
interactive. It was not designed to be used by a single person. In
short, it was hardly suited for cultural creation.
As a result of a number of developments of the 1980s and
1990sthe rise of the personal computer industry, adoption
of Graphical User Interfaces (GUI), the expansion of computer
9781623568177_txt_print.indd 55 29/04/2013 11:50
networks and the World Wide Webcomputers moved into the
cultural mainstream. Software replaced many other tools and
technologies for creative professionals. It has given hundreds of
millions of people the abilities to create, manipulate, sequence
and share mediabut has this led to the invention of fundamen-
tally new forms of culture? Today media companies are busy
promoting e-books and interactive television; the consumers are
happily purchasing music albums and feature lms distributed
in digital form, as well making photographs and video with their
digital cameras and cell phones; ofce workers are reading PDF
documents which imitate paper.
In short, it appears that the revolution in the means of production,
distribution, and access of media has not been accompanied by a
similar revolution in the syntax and semantics of media. Who
shall we blame for this? Shall we put the blame on the pioneers
of cultural computingJ. C. R. Licklider, Ivan Sutherland, Ted
Nelson, Douglas Engelbart, Seymour Paper, Nicholas Negroponte,
Alan Kay, and others? Or, as Nelson and Kay themselves are eager
to point out, does the problem lie with the way the industry imple-
mented their ideas?
Before we blame the industry for bad implementationwe can
always pursue this argument later if necessarylet us look into the
thinking of the inventors of cultural computing themselves. For
instance, what about the person who guided the development of a
prototype of a modern person computerAlan Kay?
Between 1970 and 1981 Alan Kay was working at Xerox
PARCa research center established by Xerox in Palo Alto.
Building on the already accomplished work of the pioneers of
cultural computing, the Learning Research Group at Xerox PARC
headed by Kay, systematically articulated the paradigm and the
technologies of vernacular media computing, as it exists today.
Kay has expressed his ideas in a few articles and a large number of interviews
and public lectures. The following have been my main primary sources: Alan Kay
and Adele Goldberg, Personal Dynamic Media, IEEE Computer 10, no. 3 (1977);
Alan Kay, The Early History of Smalltalk, The 2nd ACM SIGPLAN Conference
on History of Programming Languages (New York: ACM, 1993), pp. 6995; Alan
Kay, A Personal Computer for Children of All Ages, Proceedings of the ACM
1972 National Conference (Boston, 1972); Alan Kay, Doing with Images Makes
Symbols, videotape (University Video Communications, 1987), http://archive.org/
details/AlanKeyD1987/; Alan Kay, User Interface: A Personal View, in The Art of
9781623568177_txt_print.indd 56 29/04/2013 11:50
Although selected artists, lmmakers, musicians, and architects
were already using computers since the 1950s, often developing
their software in collaboration with computer scientists working
in research labs (Bell Labs, IBM Watson Research Center, etc.)
most of this software was aimed at producing only particular
kinds of images, animations or music, congruent with the ideas of
their authors. In addition, each program was designed to run on a
particular machine. Therefore, these software programs could not
function as general-purpose tools easily usable by others.
It is well known most of the key ingredients of personal computers
as they exist today came out of Xerox PARC: the Graphical User
Interface with overlapping windows and icons, bitmapped display,
color graphics, networking via Ethernet, mouse, laser printer, and
WYSIWYG (what you see is what you get) printing. But what
is equally important is that Kay and his colleagues also developed
a range of applications for media manipulation and creation that
also all used a graphical interface. They included a word processor,
a le system, a drawing and painting program, an animation
program, a music editing program, etc. Both the general user
interface and the media manipulation programs were written in the
same programming language, Smalltalk. While some of the appli-
cations were programmed by members of Kays group, the users
that included seventh-grade high-school students programmed
(This was consistent with the essence of Kays vision:
to provide users with a programming environment, examples of
programs, and already-written general tools so the users would be
able to make their own creative tools.)
When Apple introduced the rst Macintosh computer in 1984,
it brought the vision developed at Xerox PARC to consumers
(the new computer was priced at USD $2,495). The original
Macintosh 128K included a word processing and a drawing
application (MacWrite and MacPaint, respectively). Within a few
years these were joined by other software for creating and editing
Human-Computer Interface Design, ed. Brenda Laurel (Reading, Mass: Addison-
Wesley, 1990), pp. 191207; David Caneld Smith et al., Designing the Star user
Interface, Byte, issue 4 (1982).
Alan Kay and Adele Goldberg, Personal Dynamic Media, in New Media
Reader, ed. Noah Wardrip-Fruin and Nick Montfort (The MIT Press, 2003), p. 399.
9781623568177_txt_print.indd 57 29/04/2013 11:50
diferent media: Word, PageMaker and VideoWorks (1985),

SoundEdit (1986), Freehand and Illustrator (1987), Photoshop
(1990), Premiere (1991), After Efects (1993), and so on. In the
early 1990s, similar functionality became available on PCs running
Microsoft Windows.
And while Macs and PCs were at rst not
fast enough to ofer true competition for traditional media tools
and technologies (with the exception of word processing), other
computer systems specically optimized for media processing
started to compete with these technologies in the 1980s. (The
examples are the NeXT Workstation, produced between 1989 and
1996; Amiga, produced between 1985 and 1994; and Paintbox,
rst released in 1981.)
By around 1991, the new identity of a computer as a personal
media editor was rmly established. (This year Apple released
QuickTime, which brought video to the desktop; the same year
saw the release of James Camerons Terminator II, which featured
pioneering computer-generated special efects). The vision developed
at Xerox PARC became a realityor rather, one important part
of this vision in which the computer was turned into a personal
machine for display, authoring and editing content in diferent
media. And while in most cases Alan Kay and his collaborators
were not the rst to develop particular kinds of media applica-
tionsfor instance, paint programs and animation programs were
already written in the second part of the 1960s
by implementing
all of them on a single machine and giving them consistent
appearance and behavior, Xerox PARC researchers established a
new paradigm of media computing.
I think that I have made my case. The evidence is overwhelming. It
is Alan Kay and his collaborators at PARC that we must take to task
for making digital computers imitate older media. By developing
easy-to-use GUI-based software to create and edit familiar media
types, Kay and others appear to have locked the computer into
being a simulation machine for old media. Or, to put this in terms
of Jay Bolter and Richard Grusins inuential book Remediation:
Understanding New Media (2000), we can say that GUI-based
software turned a digital computer into a remediation machine:
Videoworks was renamed Director in 1987.
1982: AutoCAD; 1989: Illustrator; 1992: Photoshop, QuarkXPress.
See http://sophia.javeriana.edu.co/~ochavarr/computer_graphics_history/historia/
9781623568177_txt_print.indd 58 29/04/2013 11:50
a machine that expertly represents a range of earlier media. (Other
technologies developed at PARC, such as the bitmapped color
display used as the main computer screen, laser printing, and the
rst Page Description Language which eventually lead to Postscript,
were similarly conceived to support the computers new role as a
machine for simulation of physical media.)
Bolter and Grusin dene remediation as the representation of
one medium in another.
According to their argument, new media
always remediate the old ones and therefore we should not expect
that computers would function any diferently. This perspective
emphasizes the continuity between computational media and
earlier media. Rather than being separated by diferent logics, all
media including computers follow the same logic of remediation.
The only diference between computers and other media lies in
how and what they remediate. As Bolter and Grusin put this in
the rst chapter of their book, What is new about digital media
lies in their particular strategies for remediating television, lm,
photography, and painting. In another place in the same chapter
they make an equally strong statement that leaves no ambiguity
about their position: We will argue that remediation is a dening
characteristic of the new digital media.
If today we consider all the digital media created by both
consumers and by professionalsdigital photography and video
shot with inexpensive cameras and cell phones, the contents
of personal blogs and online journals, illustrations created
in Photoshop, feature lms cut on Avid, etc.in terms of its
appearance digital media indeed often looks exactly the same
way as media before computers. Thus, if we limit ourselves to
looking at the media surfaces, the remediation argument accurately
describes much of computational media. But rather than accepting
this condition as an inevitable consequence of the universal logic of
remediation, we should ask why this is the case. In other words, if
contemporary computational media imitates other media, how did
this become possible? There was denitely nothing in the original
theoretical formulations of digital computers by Turing or Von
Neumann about computers imitating other media such as books,
photography, or lm.
Jay Bolter and Richard Grusin, Remediation: Understanding New Media (The
MIT Press, 2000).
9781623568177_txt_print.indd 59 29/04/2013 11:50
The conceptual and technical gap which separates the rst
room-sized computers used by the military to calculate the
shooting tables for anti-aircraft guns and crack German communi-
cation codes, and contemporary small desktops and laptops used
by ordinary people to create, edit and share media is vast. The
contemporary identity of a computer as a media processor took
about forty years to emerge, if we count from 1949 when MITs
Lincoln Laboratory started to work on rst interactive computers
to 1989 when the rst commercial version of Photoshop was
released. It took generations of brilliant and creative thinkers to
invent the multitude of concepts and techniques that today make
possible for computers to remediate other media so well. What
were their reasons for doing this? What was their thinking? In
short, why did these people dedicate their careers to inventing the
ultimate remediation machine?
While media theorists have spent considerable eforts in trying
to understand the relationships between digital media and older
physical and electronic media in the 1990s and 2000s, the important
sourcesthe writing and projects by Ivan Sutherland, Douglas
Engelbart, Ted Nelson, Alan Kay, and other pioneers working in
the 1960s and 1970sremained largely unexamined. This book
does not aim to provide a comprehensive intellectual history of the
invention of media computing. Thus, I am not going to consider
the thinking of all key gures in the history of media computing
(to do this right would require more than one book). Rather, my
concern is with the present and the future. Specically, I want to
understand some of the dramatic transformations in what media
is, what it can do, and how we use itthe transformations that are
clearly connected to the shift from previous media technologies to
software. Some of these transformations had already taken place
in the 1990s but were not much discussed at the time (for instance,
the emergence of a new language of moving images and visual
design in general). Others have not even been named yet. Still
otherssuch as remix and mashup cultureare being referred to
all the time, and yet the analysis of how they were made possible
by the evolution of media software has so far not been attempted.
In short, I want to understand what is media after software
that is, what happened to the techniques, languages, and the
concepts of twentieth-century media as a result of their computer-
ization. Or, more precisely, what has happened to media after
9781623568177_txt_print.indd 60 29/04/2013 11:50
they have been software-sized. (And since in the space of a single
book I can only consider some of these techniques, languages and
concepts, I will focus on those that, in my opinion, have not been
yet discussed by others.)
In this chapter we will take a closer look at one place where the
identity of a computer as a remediation machine was largely put
in placeAlan Kays Learning Research Group at Xerox PARC,
in operation during the 1970s. We can ask two questions: rst,
what exactly did Kay want to do, and second, how did he and his
colleagues go about achieving it? The brief answerwhich will
be expanded belowis that Kay wanted to turn computers into
a personal dynamic media which could be used for learning,
discovery, and artistic creation. His group achieved this by system-
atically simulating most existing media within a computer while
simultaneously adding many new properties to these media. Kay
and his collaborators also developed a new type of programming
language that, at least in theory, would allow the users to quickly
invent new types of media using the set of general tools already
provided for them. All these tools and simulations of already
existing media were given a unied user interface designed to
activate multiple mentalities and ways of learningkinesthetic,
iconic, and symbolic.
Kay conceived of personal dynamic media as a fundamentally
new kind of media with a number of historically unprecedented
properties such as the ability to hold all the users information,
simulate all types of media within a single machine, and involve the
learner in a two-way conversation.
These properties enable new
relationships between the user and the media s/he may be creating,
editing, or viewing on a computer. And this is essential if we want to
understand the relationships between computers and earlier media.
Briey put, while visually, computational media may closely mimic
other media, these media now function in diferent ways.
For instance, consider digital photography, which often imitates
traditional photography in appearance. For Bolter and Grusin, this
is an example of how digital media remediates its predecessors.
Since the work of Kays group in the 1970s, computer scientists, hackers and
designers added many other unique propertiesfor instance, we can quickly move
media around the net and share it with millions of people using Flickr, YouTube,
and other sites.
9781623568177_txt_print.indd 61 29/04/2013 11:50
But rather than only paying attention to their appearance, let us
think about how digital photographs can function. If a digital
photograph is turned into a physical object in the worldan
illustration in a magazine, a poster on the wall, a print on a
t-shirtit functions in the same ways as its predecessor (unless
it has augmented reality features, like IKEAs 2013 catalog).

But if we leave the same photograph inside its native computer
environmentwhich may be a laptop, a network storage system,
or any computer-enabled media device such as a cell phone which
allows its user to edit this photograph and move it to other devices
and the Internetit can function in ways which, in my view,
make it radically diferent from its traditional equivalent. To use
a diferent term, we can say that a digital photograph ofers its
users many afordances that its non-digital predecessor did not.
For example, a digital photograph can be quickly modied in
numerous ways and equally quickly combined with other images;
instantly moved around the world and shared with other people;
and inserted into a multimedia document, or an architectural 3D
design. Furthermore, we can automatically (i.e., by running the
appropriate algorithms) improve its contrast, make it sharper, and
even in some situations remove blur.
Note that only some of these new properties are specic to a
particular mediumin our example, a digital photograph, i.e. an
array of pixels represented as numbers. Other properties are shared
by a larger class of media speciesfor instance, at the current stage
of digital culture, all types of media les can be attached to an email
message. Still others are even more general features of a computer
environment within the current GUI paradigm as developed forty
years ago at PARC: for instance, the fast response of a computer to
a users actions which assures no discernible pause between cause
and efect.
Still others are enabled by network protocols such
as TCP/IP that allow all kinds of computers and other devices to
be connected to the same network. In summary, we can say that
only some of the new DNA of a digital photograph is due its
particular place of birth, i.e., inside a digital camera. Many others
Roberto Baldwin, Ikeas Augmented Reality Catalog Will Let You Peek
Inside Furniture, July 20, 2012, http://www.wired.com/gadgetlab/2012/07/
Kay and Goldberg, Personal Dynamic Media, p. 394.
9781623568177_txt_print.indd 62 29/04/2013 11:50
are the result of the current paradigm of network computing in
Before diving further into Kays ideas, I should more fully
disclose my reasons for focusing on him as opposed to somebody
else. The story I will present could also be told diferently. It is
possible to put Sutherlands work on Sketchpad in the center
of computational media history; or Engelbart and his Research
Center for Augmenting Human Intellect which throughout the
1960s developed hypertext (independently of Nelson), the mouse,
the window, the word processor, mixed text/graphics displays, and
a number of other rsts. Or we can shift focus to the work of
the Architecture Machine Group at MIT, which since 1967 was
headed by Nicholas Negroponte (in 1985 this group became the
MIT Media Lab). We also need to recall that by the time Kays
Learning Research Group at PARC eshed out the details of GUI
and programmed various media editors in Smalltalk (a paint
program, an illustration program, an animation program, etc.),
artists, lmmakers and architects were already using computers
for more than a decade and a number of large-scale exhibitions of
computer art were put in major museums around the world such as
the Institute of Contemporary Art, London, The Jewish Museum,
New York, and Los Angeles County Museum of Art. And certainly,
in terms of advancing computer techniques for visual represen-
tation enabled by computers, other groups of computer scientists
were already ahead. For instance, at the University of Utah, which
became the main place for computer graphics research during the
rst half of the 1970s, scientists were producing 3D computer
graphics far superior to the simple images that could be created on
computers being built at PARC. Next to the University of Utah,
a company called Evans and Sutherland (headed by the same
Ivan Sutherland who was also teaching at the University of Utah)
was already using 3D graphics for ight simulatorsessentially
pioneering the type of new media that can be called navigable 3D
virtual space.
While the practical work accomplished at Xerox PARC to
establish the computer as a comprehensive media machine is one
of my reasons, it is not the only one. The key reason I decided to
focus on Kay is his theoretical formulations that place computers in
relation to other media and media history. While Vannevar Bush,
J. C. R. Licklider and Douglas Engelbart were primary concerned
9781623568177_txt_print.indd 63 29/04/2013 11:50
with augmentation of intellectual and in particular scientic
work, Kay was equally interested in computers as a medium of
expression through drawing, painting, animating pictures, and
composing and generating music.
Therefore if we really want
to understand how and why computers were redened as a culture
machine, and how the new computational media is diferent from
earlier physical and electronic media, I think that Kay provides us
with the best theoretical perspective.
Simulation is the central notion of
the Dynabook
While Alan Kay articulated his ideas in a number of articles and
talks, his 1977 article co-authored with one of his main PARC
collaborators, computer scientist Adele Goldberg, is a particularly
useful resource if we want to understand contemporary computa-
tional media. In this article Kay and Goldberg describe the vision
of the Learning Research Group at PARC in the following way:
to create a personal dynamic medium the size of a notebook (the
Dynabook) which could be owned by everyone and could have
the power to handle virtually all of its owners information-related
(The actual Alto computer built at Xerox PARC was the
size of later PCs; the article strategically refers to it as interim
dynabook.) Kay and Goldberg ask the readers to imagine that
this device had enough power to outrace your senses of sight
and hearing, enough capacity to store for later retrieval thousands
of page-equivalents of reference materials, poems, letters, recipes,
records, drawings, animations, musical scores, waveforms, dynamic
simulations and anything else you would like to remember and

In my view, all in the rst statement is important: it means
that the Dynabookor computational media environment in
general, regardless of the size of a form of device in which it
Ibid., p. 393.
Ibid., p. 393. The emphasis in this and all following quotes from this article is
mineL. M.
Ibid., p. 394.
9781623568177_txt_print.indd 64 29/04/2013 11:50
is implementedshould support viewing, creating and editing
all possible media traditionally used for human expression and
communication. Accordingly, while separate programs to create
works in diferent media were already in existence, Kays group
for the rst time implemented them all together within a single
machine. In other words, Kays paradigm was not to simply
create a new type of computer-based media that would co-exist
with other physical media. Rather, the goal was to establish a
computer as an umbrella, a platform for all existing expressive
artistic media. (At the end of the article Kay and Goldberg give a
name for this platform, calling it a metamedium.) This paradigm
changes our understanding of what media is. From Gotthold
Ephraim Lessings Laocoon; or, On the Limits of Painting and
Poetry (1766) to Nelson Goodmans Languages of Art (1968), the
modern discourse about media depends on the assumption that
diferent mediums have distinct properties and in fact should be
understood in opposition to each other. Putting all mediums within
a single computer environment does not necessarily erase all difer-
ences in what various mediums can represent and how they are
perceivedbut it does bring them closer to each other in a number
of ways. Some of these new connections were already apparent to
Kay and his colleagues; others became visible only decades later
when the new logic of media set in place at PARC unfolded more
fully; some may still not be visible to us today because they have
not been given practical realization. One obvious example of such
connections is the emergence of multimedia as a standard form of
communication: web pages, PowerPoint presentations, multimedia
artwork, mobile multimedia messages, media blogs, and other
communication forms which combine multiple mediums. Another
is the adoption of common interface conventions and tools which
we use in working with diferent types of media regardless of their
origin: for instance, a virtual camera, a magnifying lens, and of
course the omnipresent copy, cut and paste commands. Yet another
is the ability to map one media into another using appropriate
softwareimages into sound, sound into images, quantitative data
into a 3D shape or sound, etc.used widely today in such areas
as DJ/VJ/live cinema performances and information visualization.
All in all, it is as though diferent media are actively trying to reach
towards each other, exchanging properties and letting each other
borrow their unique features. (This situation is the direct opposite
9781623568177_txt_print.indd 65 29/04/2013 11:50
Kids learning to use the interim Dynabook. (The original caption from
the article.)
9781623568177_txt_print.indd 66 29/04/2013 11:50
The interim Dynabook system consists of processor, disk drive, display,
keyboard, and pointing devices. (The original caption from the article.)
9781623568177_txt_print.indd 67 29/04/2013 11:50
The Alto Screen showing windows with graphics drawn using commands
in Smalltalk programming language.
9781623568177_txt_print.indd 68 29/04/2013 11:50
Top: An electronic circuit layout system programmed by a 15-year-
old student Bottom: Data for this score was captured on a musical
keyboard. A program then converts the data to standard musical
notation. (The original captions from the article.)
9781623568177_txt_print.indd 69 29/04/2013 11:50
of the modernist media paradigm of the early twentieth century,
which was focused on discovering a unique language for each
artistic medium.)
Alan Turing theoretically dened a computer as a machine that
can simulate a very large class of other machines, and it is this
simulation ability that is largely responsible for the proliferation
of computers in modern society. But as I have already mentioned,
neither he nor other theorists and inventors of digital computers
explicitly considered that this simulation could also include media.
It was only Kay and his generation that extended the idea of
simulation to mediathus turning Universal Turing Machine into
a Universal Media Machine, so to speak.
Accordingly, Kay and Goldberg write: In a very real sense,
simulation is the central notion of the Dynabook.
When we
use computers to simulate some process in the real worldthe
behavior of a weather system, the processing of information in
the brain, the deformation of a car in a crashour concern is to
correctly model the necessary features of this process or system. We
want to be able to test how our model would behave in diferent
conditions with diferent data, and the last thing we want to do is
for computers to introduce some new properties into the model
that we ourselves did not specify. In short, when we use computers
as a general-purpose medium for simulation, we want this medium
to be completely transparent.
But what happens when we simulate diferent media in a
computer? In this case, the appearance of new properties may
be welcome as they can extend the expressive and communi-
cation potential of these media. Appropriately, when Kay and
his colleagues created computer simulations of existing physical
mediai.e. the tools for representing, creating, editing, and viewing
these mediathey added many new properties. For instance, in
the case of a book, Kay and Goldberg point out It need not be
treated as a simulated paper book since this is a new medium with
new properties. A dynamic search may be made for a particular
context. The non-sequential nature of the le medium and the use
of dynamic manipulation allow a story to have many accessible
points of view.
Kay and his colleagues also added various other
Ibid., p. 399.
Ibid., p. 395.
9781623568177_txt_print.indd 70 29/04/2013 11:50
properties to the computer simulation of paper documents. As Kay
has referred to this in another article, his idea was not to simply
imitate paper but rather to create magical paper.
For instance,
the PARC team gave users the ability to modify the fonts in a
document and create new fonts. They also implemented another
important idea that had already been developed by Douglas
Engelbarts team in the 1960s: the ability to create diferent views
of the same structure (I will discuss this in more detail below). And
both Engelbart and Ted Nelson had already added something
else: the ability to connect diferent documents or diferent parts of
the same document through hyperlinkingi.e. what we now know
as hypertext and hypermedia. Engelbarts group also developed
the ability for multiple users to collaborate on the same document.
This list goes on and on: e-mail in 1965, newsgroups in 1979,
World Wide Web in 1990, etc.
Each of these new properties had far-reaching consequences.
Take Search, for instance. Although the ability to search through
a page-long text document does not sound like a very radical
innovation, as the document gets longer this ability becomes more
and more important. It becomes absolutely crucial if we have a
very large collection of documentssuch as all the web pages
on the Web. Although current search engines are far from being
perfect and new technologies will continue to evolve, imagine how
diferent the culture of the Web would be without them.
Or take the capacity to collaborate on the same document(s) by
a number of users connected to the same network. While it was
already widely used by companies in the 1980s and 1990s, it was
not until the early 2000s that the wider public saw the real cultural
potential of this addition to print media. By harvesting the small
amounts of labor and expertise contributed by a large number of
volunteers, social software projectsmost famously, Wikipedia
created vast and dynamically updatable pools of knowledge which
would be impossible to create in traditional ways. (In a less visible
way, every time we do a search on the Web and then click on
some of the results, we also contribute to a knowledge-set used by
everybody else. In deciding in which sequence to present the results
of a particular search, Googles algorithms take into account which
Alan Kay, User Interface: A Personal View, p. 199.
9781623568177_txt_print.indd 71 29/04/2013 11:50
among the results of previous searches for the same words people
found most useful.)
Studying the writings and public presentations of the people
who invented interactive media computingSutherland, Engelbart,
Nelson, Negroponte, Kay, and othersmakes it clear that they did
not produce the new properties of computational media as an after-
thought. On the contrary, they knew that they were turning physical
media into new media. In 1968 Engelbart gave his famous demo at
the Fall Joint Computer Conference in San Francisco before a few
thousand people that included computer scientists, IBM engineers,
people from other companies involved in computers, and funding
ofcers from various government agencies.
Although Engelbart had
only ninety minutes, he had a lot to show. Over the few previous years,
his team at The Research Center for Augmenting Human Intellect
had essentially developed the modern ofce environment as it exists
today (not be confused with the modern media design environment
which was developed later at PARC). Their NLS computer system
included word processing with outlining features, documents
connected through hypertext, online collaboration (two people at
remote locations working on the same document in real-time), online
user manuals, online project planning systems, and other elements of
what is now called computer-supported collaborative work. The
team also developed the key elements of modern user interface that
were later rened at PARC: a mouse and multiple windows.
Paying attention to the sequence of the demo reveals that while
Engelbart had to make sure that his audience would be able to relate
the new computer system to what they already knew and used, his
focus was on new features of simulated media never before available
Engelbart devotes the rst segment of the demo to
word processing, but as soon as he briey demonstrated text entry,
cut, paste, insert, naming and saving lesin other words, the set
of tools which make a computer into a more versatile typewriter
M. Mitchell Waldrop, The Dream Machine: J. C. R. Licklider and the Revolution
That Made Computing Personal (Viking, 2001), p. 287.
Complete video of Engelbardts 1968 demo is available at http://sloan.stanford.
edu/MouseSite/1968Demo.html. For the detailed descriptions of NLS functions, see
Augmentation Research Center, NLS User Training Guide, Stanford Research
Institute: Menlo Park, California), 1997, http://bitsavers.org/pdf/sri/arc/NLS_User_
9781623568177_txt_print.indd 72 29/04/2013 11:50
he then goes on to show in more depth the features of his system
which no writing medium had before: view control. As Engelbart
points out, the new writing medium could switch at the users wish
between many diferent views of the same information. A text le
could be sorted in diferent ways. It could also be organized as
a hierarchy with a number of levels, as in outline processors or
outlining mode of contemporary word processors such as Microsoft
Word. For example, a list of items can be organized by categories
and individual categories can be collapsed and expanded.
Engelbart next shows another example of view control, which
today, forty-ve years after his demo, is still not available in popular
document management software. He makes a long to do list and
organizes it by locations. He then instructs the computer to display
these locations as a visual graph (a set of points connected by
lines.) In front of our eyes, representation in one medium changes
into another mediumtext becomes a graph. But this is not all.
The user can control this graph to display diferent amounts of
informationsomething that no image in physical media can do.
As Engelbart clicks on diferent points in a graph corresponding
to particular locations, the graph shows the appropriate part of
his to do list. (This ability to interactively change how much
and what information an image shows is particularly important in
todays information visualization applications.)
Next Engelbart presents a chain of views which he prepared
beforehand. He switches between these views using links which
may look like hyperlinks the way they exist on the Web todaybut
they actually have a diferent function. Instead of creating a path
between many diferent documents la Vannevar Bushs Memex
(often seen as the precursor to modern hypertext), Engelbart is
using links as a method for switching between diferent views of a
single document organized hierarchically. He brings a line of words
displayed in the upper part of the screen; when he clicks on these
words, more detailed information is displayed in the lower part
of the screen. This information can in turn contain links to other
views that show even more detail.

For the detailed descriptions of these and other capabilities of NLS, see
Augmentation Research Center, NLS User Training Guide, Stanford Research
Institute: Menlo Park, California), 1997, http://bitsavers.org/pdf/sri/arc/NLS_User_
9781623568177_txt_print.indd 73 29/04/2013 11:50
Examples of view control as implemented in NLS. Top: a hierarchical
view of a shopping list. Top: a collapsed view sorted by location. Bottom:
a graph view showing the sequence of locations. (Text and graphics were
traced from the original video of Engelbarts 1968 demo.)
9781623568177_txt_print.indd 74 29/04/2013 11:50
Rather than using links to drift through the textual universe
associatively and horizontally, we move vertically between
more general and more detailed information. Appropriately, in
Engelbarts paradigm, we are not navigatingwe are switching
views. We can create many diferent views of the same infor-
mation and switch between these views in diferent ways. And
this is what Engelbart systematically explains in this rst part of
his demo. He demonstrates that you can change views by issuing
commands, by typing numbers that correspond to diferent parts of
a hierarchy, by clicking on parts of a picture, or on links in the text.
(In 1967 Ted Nelson articulated, named a similar idea of a type of
hypertext, which would allow a reader to obtain a greater detail
on a specic subject. He named it stretchtext.
Since new media theory and criticism emerged in the early
1990s, endless texts have been written about interactivity,
hypertext, virtual reality, cyberspace, cyberculture, cyborgs, and
so on. But I have never seen anybody discuss view control. And
yet this is one of the most fundamental and radical new techniques
for working with information and media available to us today. It
is used daily by each of us numerous times. View control, i.e.
the abilities to switch between many diferent views and kinds of
views of the same information is now implemented in multiple
ways not only in OS, word processors and email clients, but also
in all media processors (i.e. media editing software): AutoCAD,
Maya, After Efects, Final Cut, Photoshop, InDesign, and so on.
For instance, in the case of 3D software, it can usually display the
model in at least half a dozen diferent ways: in wireframe, fully
rendered, etc. In the case of animation and visual efects software,
since a typical project may contain dozens of separate objects each
having dozens of parameters, it is often displayed in a way similar
to how outline processors can show text. In other words, the user
can switch between more and less information. You can choose
to see only those parameters which you are working on right
now. You can also zoom in and out of the composition. When
you do this, parts of the composition do not simply get smaller
or biggerthey show less or more information automatically. For
instance, at a certain scale you may only see the names of diferent
Ted Nelson, Stretchtext (Hypertext Note 8), 1967, http://xanadu.com/
9781623568177_txt_print.indd 75 29/04/2013 11:50
View control as implemented in Macintosh System software, 1984. Top:
applications, folders, and les in Guided Tour oppy disk. Bottom:
View of applications, folders, and les sorted by icon.
9781623568177_txt_print.indd 76 29/04/2013 11:50
Top: View of applications, folders, and les sorted by date. Bottom: View
of applications, folders, and les sorted by size.
9781623568177_txt_print.indd 77 29/04/2013 11:50
parameters; but when you zoom into the display, the program
may also display the graphs which indicate how these parameters
change over time.
Let us look at another exampleTed Nelsons concept of
hypertext that he developed in the early 1960s (independently but
parallel to Engelbart).
In his 1965 article A File Structure for the
Complex, the Changing, and the Indeterminate, Nelson discusses
the limitations of books and other paper-based systems for organ-
izing information and then introduces his new concept:
However, with the computer-driven display and mass memory,
it has become possible to create a new, readable medium, for
education and enjoyment, that will let the reader nd his level,
suit his taste, and nd the parts that take on special meaning for
him, as instruction and enjoyment.
Let me introduce the word hypertext to mean a body of
written or pictorial material interconnected in such a complex
way that it could not be conveniently presented or represented
on paper.
A new, readable mediumthese words make it clear that Nelson
was not simply interested in patching up books and other paper
documents. Instead, he wanted to create something distinctively
new. But was not hypertext as proposed by Nelson simply an
extension of older textual practices such as exegesis (extensive
interpretations of holy scriptures such as the Bible, Talmud,
Qurn), annotations, or footnotes? While such historical prece-
dents for hypertext are often proposed, they mistakenly equate
Nelsons proposal with a very limited form in which hypertext is
experienced by most people todayi.e., the World Wide Web. As
Noah Wardrip-Fruin pointed out, The Web implemented only
Douglas C. Engelbart, Augmenting Human Intellect: A Conceptual Framework
(Stanford Research Institute, 1962), http://www.dougengelbart.org/pubs/augment-
3906.html. Although the implementation of hypertext in Engelbarts NLS was much
more limited than Nelsons concept of hypertext, looking at Engelbarts discussion
in Augmenting Human Intellect shows that his ideas for new systems for organizing
information were at least as rich as Nelsons.
Theodor H. Nelson, A File Structure for the Complex, the Changing, and the
Indeterminate (1965), in New Media Reader, p. 144.
9781623568177_txt_print.indd 78 29/04/2013 11:50
one of many types of structures proposed by Nelson already in
1965chunk style hypertextstatic links that allow the user to
jump from page to page.

Following the Web implementation, most people today think
of hypertext as a body of text connected through one-directional
links. However, the terms links does not even appear in Nelsons
original denition of hypertext. Instead, Nelson talks about new
complex interconnectivity without specifying any particular mecha-
nisms that can be employed to achieve it. A particular system
proposed in Nelsons 1965 article is one way to implement such
a vision, but as his denition implicitly suggests, many others are
also possible.
What kind of structures are possible in hypertext? asks
Nelson in a research note from 1967. He answers his own question
in a short but very suggestive manner: Any.
Nelson goes on
to explain: Ordinary text may be regarded as a special casethe
simple and familiar caseof hypertext, just as three-dimensional
space and the ordinary cube are the simple and familiar special
cases of hyperspace and hypercube.
(In 2007 Nelson re-stated
this idea in the following way: Hypertexta word I coined long
agois not technology but potentially the fullest generalization of
documents and literature.
If hypertext does not simply mean links, it also does not
only mean text. Although in its later popular use the word
hypertext came to refer to linked text, as one can see from the
quote above, Nelson included pictures in his denition of hyper-
And in the following paragraph, he introduces the terms
hyperlm and hypermedia:
Noah Wardrip-Fruin, introduction to Theodor H. Nelson, A File Structure for
the Complex, the Changing, and the Indeterminate (1965), in New Media Reader,
p. 133.
Ted Nelson, Brief Words on the Hypertext (Hypertext Note 1), 1967, http://
Ted Nelson, http://transliterature.org/ (version TransHum-D23, 07.06.17).
In his presentation at the 2004 Digital Retroaction symposium Noah Wardrip-
Fruin stressed that Nelsons vision included hypermedia and not only hypertext.
Noah Wardrip-Fruin, presentation at Digital Retroaction: a Research Symposium,
UC Santa Barbara, September 1719, 2005, http://dc-mrg.english.ucsb.edu/
9781623568177_txt_print.indd 79 29/04/2013 11:50
Films, sound recordings, and video recordings are also linear
strings, basically for mechanical reasons. But these, too, can now
be arranged as non-linear systems for instance, lattices for
educational purposes, or for display with diferent emphasis
The hyperlm a browsable or vari-sequenced movie is only
one of the possible hypermedia that require our attention.
Where is hyperlm today, almost 50 years after Nelson articulated
this concept? If we understand hyperlm in the same limited sense
as hypertext is understood todayshots connected through links
which a user can click onit would seems that hyperlm never
fully took of. A number of early pioneering projectsAspen
Movie Map (Architecture Machine Group, 19789), Earl King
and Sonata (Grahame Weinbren, 19835; 19913), CD-ROMs
by Bob Steins Voyager Company, and Wax: Or the Discovery
of Television Among the Bees (David Blair, 1993)have not
been followed up. Similarly, interactive movies and FMV-games
created by the video game industry in the rst half of the 1990s
soon fell out of favor, replaced by 3D games (which ofered more
interactivity). But if instead we think of hyperlm in a broader
sense, as it was conceived by Nelsonany interactive structure
for connecting video or lm elements, with a traditional lm being
a special casewe realize that hyperlm is much more common
today than it may appear. Numerous interactive Flash and HTML5
sites which use video, video clips with markers which allow a user
jump to a particular point in a video (for instance, see the videos on
), and database cinema
are just some of the examples
of hyperlm today.
Decades before hypertext and hypermedia became the common
ways for interacting with information, Nelson understood well
what these ideas meant for our well-established cultural practices
and concepts. The announcement for his January 5, 1965 lecture at
Vassar College talks about this in terms that are even more relevant
today than they were then: The philosophical consequences of all
this are very grave. Our concepts of reading, writing, and book
fall apart, and we are challenged to design hyperles and write
Nelson, A File Structure, p. 144.
www.ted.com (March 8, 2008).
See http://softcinema.net/form.htm
9781623568177_txt_print.indd 80 29/04/2013 11:50
hypertext that may have more teaching power than anything that
could ever be printed on paper.
These statements align Nelsons thinking and work with artists
and theorists who similarly wanted to destabilize the conventions
of cultural communication. Digital media scholars extensively
discussed parallels between Nelson and French theorists writing
during the 1960sRoland Barthes, Michel Foucault and Jacque
Others have pointed out close parallels between the
thinking of Nelson and literary experiments taking place around
the same time, such as works by Oulipo.
(We can also note
the connection between Nelsons hypertext and the non-linear
structure of the lms of French lmmakers who set out to question
the classical narrative style: Hiroshima Mon Amour, Last Year at
Marienbad, Breathless and others).
How far shall we take these parallels? In 1987 Jay Bolter and
Michael Joyce wrote that hypertext could be seen as a continu-
ation of the modern tradition of experimental literature in print
which includes modernism, futurism, Dada surrealism, lettrism,
the nouveau roman, concrete poetry.
Refuting their claim, Espen
J. Aarseth has argued that hypertext is not a modernist structure
per se, although it can support modernist poetics if the author
desires this.
Who is right? Since this book argues that cultural
software turned media into metamediaa fundamentally new
semiotic and technological system which includes most previous
media techniques and aesthetics as its elementsI also think
that hypertext is actually quite diferent from modernist literary
tradition. I agree with Aarseth that hypertext is indeed much
more general than any particular poetics such as modernist ones.
Announcement of Ted Nelsons lecture at Vassar College, January 5, 1965, http://
George Landow, ed., Hypertext: The Convergence of Contemporary Critical
Theory and Technology (The Johns Hopkins University Press, 1991); Jay Bolter,
The writing space: the computer, hypertext, and the history of writing (Hillsdale,
NJ: L. Erlbaum Associates, 1991).
Randall Packer and Ken Jordan, Multimedia: From Wagner to Virtual Reality
(W. W. Norton & Company, 2001); Noah Wardrip-Fruin and Nick Monford, New
Media Reader (The MIT Press, 2003).
Quoted in Espen J. Aarseth, Cybertext: Perspectives on Ergodic Literature (The
Johns Hopkins University Press, 1997), p. 89.
Espen J. Aarseth, Cybertext, 8990.
9781623568177_txt_print.indd 81 29/04/2013 11:50
Indeed, already in 1967 Nelson said that hypertext could support
any structure of information including that of traditional texts
and presumably, this also includes diferent modernist poetics.
(Importantly, this statement is echoed in Kay and Goldbergs
denition of the computer as a metamedium whose content is
a wide range of already-existing and not-yet-invented media.)
What about the scholars who see the strong connections
between the thinking of Nelson and modernism? Although Nelson
says that hypertext can support any information structure and that
this information does not need to be limited to text, his examples
and his style of writing show an unmistakable aesthetic sensi-
bilitythat of literary modernism. He clearly dislikes ordinary
text. The emphasis on complexity and interconnectivity and on
breaking up conventional units for organizing information such
as a page clearly aligns Nelsons proposal for hypertext with
the early twentieth-century experimental literaturethe inven-
tions of Virginia Woolf, James Joyce, the Surrealists, etc. This
connection to literature is not accidental since Nelsons original
motivation for his research that led to hypertext was to create a
system for handling both the notes for literary manuscripts and
those manuscripts themselves. Nelson also already knew about the
writings of William Burroughs. The very title of the articleA File
Structure for the Complex, the Changing, and the Indeterminate
would make the perfect title for an early twentieth-century
avant-garde manifesto, as long as we substitute le structure
with some ism.
Nelsons modernist sensibility also shows itself in his thinking
about new mediums that can be established with the help of a
computer. However, his work should not be seen as a simple
continuation of modernist tradition. Rather, both his and Kays
research represent the next stage of the avant-garde project.
The early twentieth-century avant-garde artists were primarily
interested in questioning conventions of established media such
as photography, print, graphic design, cinema, and architecture.
Thus, no matter how unconventional the paintings that came
out from Futurism, Orphism, Suprematism or De Stijl were, their
manifestos were still talking about them as paintingsrather
than as a new media. In contrast, Nelson and Kay explicitly write
about creating new media, not only changing the existing ones.
Nelson: With the computer-driven display and mass memory, it
9781623568177_txt_print.indd 82 29/04/2013 11:50
has become possible to create a new, readable medium. Kay and
Goldberg: It [computer text] need not be treated as a simulated
paper book since this is a new medium with new properties.
Another key diference between how modernist artists and
pioneers of cultural software approached the job of inventing
new media and extending existing ones is captured by the title
of Nelsons article I have been already quoting above: A File
Structure for the Complex, the Changing, and the Indeterminate.
Instead of a particular modernist ism, we get a le structure.
Cubism, Expressionism, Futurism, Orphism, Suprematism, and
Surrealism proposed new distinct systems for organizing infor-
mation, with each system ghting all others for the dominance
in the cultural memesphere. In contrast, Bush, Licklider, Nelson,
Engelbart, Kay, Negroponte, and their colleagues created meta-
systems that can support many kinds of information structures.
Kay called such a system a rst metamedium, Nelson referred to
it as hypertext and hypermedia, Engelbart wrote about automated
external symbol manipulation and bootstrapping,but behind
the diferences in their visions lay the similar understanding of
the radically new potential ofered by computers for information
manipulation. The prexes meta- and hyper- used by Kay
and Nelson were the appropriate characterizations for a system
which was more than another new medium that could remediate
other media in its particular ways. Instead, the new system would
be capable of simulating all these media with all their remediation
strategiesas well as supporting development of what Kay and
Goldberg referred to as new not-yet-invented media. And of
course, this was not all. Equally important was the role of inter-
activity. The new meta-systems proposed by Nelson, Kay and
others were to be used interactively to support the processes of
thinking, discovery, decision making, and creative expression. In
contrast, the aesthetics created by modernist movements could be
understood as information formatting systemsto be used for
selecting and organizing information into xed presentations that
are then distributed to the users, not unlike PowerPoint slides.
Finally, at least in Kays and Nelsons vision, the task of dening
new information structures and media manipulation techniques
and, in fact, new media as a wholewas given to the user, rather
than being the sole province of the designers. This decision had
far-reaching consequences for shaping contemporary culture. Once
9781623568177_txt_print.indd 83 29/04/2013 11:50
computers and programming were democratized enough, many
creative people started to focus on creating these new structures and
techniques rather than using the existing ones to make content.
Since the end of 2000, extending the computer metamedium by
writing new software, plugins, programming libraries and other
tools became the new cutting-edge type of cultural activity giving
a new meaning to McLuhans famous formula the medium is the
Today a typical article in computer science or information science
will not be talking about inventing a new medium as a justi-
cation for research. Instead, it is likely to refer to previous work
in some eld or sub-eld of computer science such as knowledge
discovery, data mining, semantic web, etc. It can also refer to
existing social and cultural practices and industriesfor instance,
e-learning, video game development, collaborative tagging,
or massively distributed collaboration. In either case, the need
for new research is justied by a reference to already established or
popular practicesacademic paradigms which have been funded,
large-scale industries, and mainstream social routines which do
not threaten or question the existing social order. This means
that practically all of computer science research which deals with
mediaweb technologies, media computing, hypermedia, human-
computer interfaces, computer graphics, and so onis oriented
towards mainstream media usage.
In other words, either computer scientists are trying to make
more efcient the technologies already used in media industries
(video games, web search engines, lm production, etc.) or they
are inventing new technologies that are likely to be used by these
industries in the future. The invention of new mediums for its own
sake is not something which anybody is likely to pursue, or get
funded. From this perspective, the software industry and business
in general is often more innovative than academic computer
science. For instance, social media applications (Wikipedia, Flickr,
YouTube, Facebook, del.icio.us, Digg, etc.) were not invented in the
academy; nor were HyperCard, QuickTime, HTML, Photoshop,
After Efects, Flash, or Google Earth. This was no diferent in
previous decades. It is, therefore, not accidental that the careers of
both Ted Nelson and Alan Kay were spent in the industry and not
the academy: Kay worked for and was a fellow at Xerox PARC,
Atari, Apple and Hewlett-Packard; Nelson was a consultant and
9781623568177_txt_print.indd 84 29/04/2013 11:50
a fellow at Bell Laboratories, Datapoint Corporation, Autodesk;
both were also associated with Disney.
Why did Nelson and Kay nd more support in industry than in
academia for their quest to invent new computer media? And why
is the industry (by which I simply mean any entity which creates
the products which can be sold in large quantities, or monetized
in other ways, regardless of whether this entity is a large multina-
tional company or a small start-up)more interested in innovative
media technologies, applications, and content than computer
science? The systematic answer to this question will require its
own investigation. Also, what kinds of innovations each modern
institution can support changes over time. But here is one brief
answer: modern business thrives on creating new markets, new
products, and new product categories. Although the actual devel-
opment of such new markets and products is always risky, it is also
very protable. This was already the case in the previous decades
when Nelson and Kay were supported by Xerox, Atari, Apple, Bell
Labs, Disney, etc. In the 2000s, following the globalization of the
1990s, all areas of business embraced innovation to an unprec-
edented degree; this pace quickened around 2005 as companies
fully focused on competing for new consumers in China, India,
and other emerging economies. Around the same time, we saw
a similar increase in the number of innovative products in the IT
industry: open APIs of leading Web 2.0 sites, daily announcements
of new web services, locative media applications, new innovative
products such as iPhone, new paradigms in imaging such as HDR
and non-destructive editing, the beginnings of a long tail for
software, open source hardware, and so on.
As we can see from the examples we have analyzed, the aim
of the inventors of computational mediaEngelbart, Nelson,
Kay and the people who worked with themwas not simply to
create accurate simulations of physical media. Instead, in every
case the goal was to create a new medium with new properties
which would allow people to communicate, learn, and create in
new ways. So while today the content of these new media may
often look the same as that of its predecessors, we should not
be fooled by this similarity. The newness lies not in the content
but in the software tools used to create, edit, view, distribute,
and share this content. Therefore, rather than only looking at the
output of software-based cultural practices, we need to consider
9781623568177_txt_print.indd 85 29/04/2013 11:50
software itselfsince it allows people to work with media in a
number of historically unprecedented ways. So while on the level
of appearance computational media indeed often remediate (i.e.
represent) previous media, the software environment in which this
media lives is very diferent.
Let me add two more examples. One is Ivan Sutherlands
Sketchpad (1962). Created by Sutherland as a part of his PhD
thesis at MIT, Sketchpad deeply inuenced all subsequent work in
computational media (including that of Kay) not only because it
was the rst interactive media authoring program but also because
it made it clear that computer simulations of physical media can
add many exciting new properties to the media being simulated.
Sketchpad was the rst software that allowed its users to interac-
tively create and modify line drawings. As Noah Wardrip-Fruin
pointed out, it moved beyond paper by allowing the user to work
at any of 2000 levels of magnicationenabling the creation of
projects that, in physical media, would either be unwieldy large
or require detail work at an impractically small size.
similarly redened graphical elements of a design as objects which
can be manipulated, constrained, instantiated, represented ironi-
cally, copied, and recursively operated upon, even recursively
For instance, if the designer dened new graphical
elements as instances of a master element and later made a change
to the master, all these instances would also change automatically.
Another new property, which perhaps demonstrated most
dramatically how computer-aided drafting and drawing were
diferent from their physical counterparts, was Sketchpads use
of constraints. In Sutherlands own words, The major feature
which distinguishes a Sketchpad drawing from a paper and pencil
drawing is the users ability to specify to Sketchpad mathematical
conditions on already drawn parts of his drawing which will be
automatically satised by the computer to make the drawing take
the exact shape desired.
For instance, if a user drew a few lines,
and then gave the appropriate command, Sketchpad automatically
Noah Wardrip-Fruin, introduction to Sketchpad. A Man-Machine Graphical
Communication System, in New Media Reader, 1963, p. 109.
Ivan Sutherland, Sketchpad. A Man-Machine Graphical Communication
System, Proceedings of the AFIPS Spring Joint Computer Conference, Detroit,
9781623568177_txt_print.indd 86 29/04/2013 11:50
Frames from Sketchpad demo video illustrating the programs use of
constraints. Left column: a user selects parts of a drawing. Right column:
Sketchpad automatically adjusts the drawing. (The captured frames were
edited in Photoshop to show the Sketchpad screen more clearly.)
9781623568177_txt_print.indd 87 29/04/2013 11:50
moved these lines until they were parallel to each other. If a user
gave a diferent command and selected a particular line, Sketchpad
moved the lines in such a way so they would parallel to each other
and perpendicular to the selected line.
Although we have not exhausted the list of new properties that
Sutherland built into Sketchpad, it should be clear that this rst
interactive graphical editor was not only simulating existing media.
Appropriately, Sutherlands 1963 paper on Sketchpad repeatedly
emphasizes the new graphical capacities of his system, marveling
how it opens new elds of graphical manipulation that has never
been available before.
The very title given by Sutherland to
his PhD thesis foregrounds the novelty of his work: Sketchpad:
A man-machine graphical communication system. Rather than
conceiving of Sketchpad as simply another medium, Sutherland
presents it as something elsea communication system between
two entities: a human and an intelligent machine. Kay and
Goldberg later also foregrounded this communication dimension,
referring to it as a two-way conversation and calling the new
metamedium active.
(We can also think of Sketchpad as a
practical demonstration of the idea of man-machine symbiosis
by J. C. R. Licklider applied to image making and design.
My last example comes from the software development that at
rst sight may appear to contradict my argument: paint software.
Surely, the applications which simulate in detail the range of
efects made possible with various physical brushes, paint knives,
canvases, and papers are driven by the desire to recreate the
experience of working within an existing medium rather than the
desire to create a new one? Wrong. In 1997 an important computer
graphics pioneer Alvy Ray Smith wrote a memo titled Digital Paint
Systems: Historical Overview.
In this text Smith (who himself
had a background in art) makes an important distinction between
Michigan, May 213, 1963, pp. 32946; in New Media Reader, Noah Wardrip-
Fruin and Nick Montfort (eds).
Ibid., p. 123.
Kay and Goldberg, Personal Dynamic Media, 394.
J. C. R. Licklider, Man-Machine Symbiosis, IRE Transactions on Human
Factors in Electronics, vol. HFE-1, March 1960, pp. 411, in New Media Reader,
eds. Noah Wardrip-Fruin and Nick Montfort.
Alvy Ray Smith, Digital Paint Systems: Historical Overview (Microsoft Technical
Memo 14, May 30, 1997). http://alvyray.com/
9781623568177_txt_print.indd 88 29/04/2013 11:50
digital paint programs and digital paint systems. In his denition,
A digital paint program does essentially no more than implement
a digital simulation of classic painting with a brush on a canvas.
A digital paint system will take the notion much farther, using
the simulation of painting as a familiar metaphor to seduce
the artist into the new digital, and perhaps forbidding, domain.
(Emphasis in the original). According to Smiths history, most
commercial painting applications, including Photoshop, fall into
the paint system category. His genealogy of paint systems begins
with Richard Shoups SuperPaint, developed at Xerox PARC in
While SuperPaint allowed the user to paint with a variety
of brushes in diferent colors, it also included many techniques not
possible with traditional painting or drawing tools. For instance, as
described by Shoup in one of his articles on SuperPaint, Objects
or areas in the picture may be scaled up or down in size, moved,
copied, overlaid, combined or changed in color, and saved on disk
for future use or erased.

Most important, however, was the ability to grab frames from
video. Once loaded into the system, such a frame could be treated
as any other imagethat is, an artist could use all of SuperPaints
drawing and manipulation tools, add text, combine it with other
images, etc. The system could also translate what appeared on its
screen back into a video signal. Accordingly, Shoup is clear that
his system was much more than a way to draw and paint with
a computer. In a 1979 article, he refers to SuperPaint as a new
videographic medium.
In another article published a year
later, he renes this claim: From a larger perspective, we realized
that the development of SuperPaint signaled the beginning of the
synergy of two of the most powerful and pervasive technologies
ever invented: digital computing and video or television.
This statement is amazingly perceptive. When Shoup was
writing this in 1980, computer graphics were used in television
Richard Shoup, SuperPaint: An Early Frame Bufer Graphics Systems, IEEE
Annals of the History of Computing 23, issue 2 (AprilJune 2001), p. 327,
http://www.rgshoup.com/prof/SuperPaint/Annals_final.pdf; Richard Shoup,
SuperPaintThe Digital Animator, Datamation (1979), http://www.rgshoup.
Shoup, SuperPaintThe Digital Animator, p. 152.
Ibid., p. 156.
Shoup, SuperPaint: An Early Frame Bufer Graphics System, p. 32.
9781623568177_txt_print.indd 89 29/04/2013 11:50
broadcasts just a handful of times. And while in the next decade
their use became more common, only in the middle of the 1990s
did the synergy Shoup predicted truly became visible. As we will
see in the chapter on After Efects below, the result was a dramatic
reconguration not just of the visual languages of television but of
all visual techniques invented by humans up to that point. In other
words, what began as a new videographic medium in 1973 had
eventually changed all visual media.
But even if we forget about SuperPaints revolutionary ability
to combine graphics and video, and discount its new tools such
resizing, moving, copying, etc., we are still dealing with a new
creative medium (Smiths term). As Smith pointed out, this medium
is the digital frame bufer,
a special kind of computer memory
Alvy Ray Smith, Digital Paint Systems: An Anecdotal and Historical Overview,
IEEE Annals of the History of Computing. 2011, http://accad.osu.edu/~waynec/
SuperPaint menu, 1975.
9781623568177_txt_print.indd 90 29/04/2013 11:50
designed to hold images represented as an array of pixels (today
a more common name is graphics card). An artist using a paint
system is modifying pixel values in a frame buferregardless of
what particular operation or tool s/he is employing at the moment.
This opens up a door to all kinds of new image creation and
modication operations, which follow diferent logic than physical
painting. The telling examples of this can be found in a paint system
called Paint developed by Smith in 19756. In Smiths own words,
Instead of just simulating painting a stroke of constant color, I
extended the notion to mean perform any image manipulation
you want under the pixels of the paintbrush.
Beginning with
this conceptual generalization, Smith added a number of efects
which still used a paintbrush tool but actually no longer referred
to painting in a physical world. For instance, in Paint any image
of any shape could be used as a brush. In another example, Smith
added not paint that reversed the color of every pixel under the
paintbrush to its color complement. He also dened smear paint
that averaged the colors in the neighborhood of each pixel under
the brush and wrote the result back into the pixel. And so on.
Thus, the instances where the paintbrush tool behaved more like a
real physical paintbrush were just particular cases of a much larger
universe of new behaviors made possible in a new medium.
The permanent extendibility
As we saw, Sutherland, Nelson, Engelbart, Kay, and other pioneers
of computational media have added many previously non existent
properties to media that they have simulated in a computer.
The subsequent generations of computer scientists, hackers, and
designers added many more propertiesbut this process is far
from nished. And there is no logical or material reason why it
will ever be nished. It is the nature of computational media
that it is open-ended and that new techniques are continuously
being invented.
To add new properties to physical media requires modifying
its physical substance. But since computational media exists as
Ibid., p. 18.
9781623568177_txt_print.indd 91 29/04/2013 11:50
software, we can add new properties or even invent new types of
media by simply changing existing or writing new software. Or by
adding plug-ins and extensions, as programmers have been doing
it with Photoshop and Firefox, respectively. Or by putting existing
software together. (For instance, starting in 2006, thousands of
people extended the capacities of mapping media by creating
software mashups which combine the services and data provided
by Goggle Maps, Flickr, Amazon, other sites, and media uploaded
by users.)
In short, new media is new because new properties (i.e.,
new software techniques) can always be easily added to it. Put
diferently, in industrial (i.e. mass-produced) media technologies,
hardware and software were one and the same thing. For
example, the book pages were bound in a particular way that xed
the order of pages. The reader could not change this order nor
the level of detail being displayed la Engelbarts view control.
Similarly, the lm projector combined hardware and what we now
call a media player software into a single machine. In the same
way, the controls built into a twentieth-century mass-produced
camera could not be modied at the users will. And although
today the users of a digital camera similarly cannot easily modify
the hardware of their camera, as soon as they transfer the pictures
into a computer they have access to endless number of controls and
options for modifying their pictures via software.
In the nineteenth and twentieth centuries the normally rigid
industrial media was uid in two situations. First, when a new
media was being rst developed: for instance, the invention of
photography in the 1820s1840s. Second, when artists would
systematically experiment with and open up already industri-
alized mediasuch as the experiments with lm and video during
the 1960s that came to be called Expanded Cinema.
What used to be separate moments of experimentations with
media during the industrial era became the norm in a software
society. In other words, the computer legitimizes experimen-
tation with media. Why is this so? What diferentiates a modern
digital computer from any other machineincluding industrial
media machines for capturing and playing mediais separation
of hardware and software. It is because an endless number of
diferent programs performing diferent tasks can be written to
run on the same type of machine, that that machinei.e. a digital
9781623568177_txt_print.indd 92 29/04/2013 11:50
computeris used so widely today. Consequently, the constant
invention of new (and modication of existing) media software, is
simply one example of this general principle. In its very structure
computational media is avant-garde since it is constantly being
extended and thus redened.
If in modern culture experimental and avant-garde were
opposed to normalized and stable, this opposition largely disap-
pears in software culture. And the role of the media avant-garde
is performed no longer by individual artists in their studios but by
a variety of players, from very big to very smallfrom companies
such as Microsoft, Adobe, and Apple to independent programmers,
hackers, and designers.
But this process of continual invention of new algorithms does
not just move in any direction. If we look at contemporary media
softwareCAD, computer drawing and painting, image editing,
word processorswe will see that most of their fundamental
principles were already developed by the generation of Sutherland
and Kay. In fact the very rst interactive graphical editor
Sketchpadalready contains most of the genes, so to speak, of
contemporary graphics applications. As new techniques continue
to be invented they are layered over the foundations that were
gradually put in place by Sutherland, Engelbart, Kay, and others in
the 1960s and 1970s.
Of course we are not dealing here only with the history of ideas.
Various social and economic factorssuch as the dominance
of the media software market by a handful of companies or the
wide adoption of particular le formats also constrain possible
directions of software evolution. Put diferently, today software
development is an industry and as such it is constantly balancing
between stability and innovation, standardization and exploration
of new possibilities. But it is not just any industry. New programs
can be written and existing programs can be extended and modied
(if the source code is available) by anybody who has programming
skills and access to a computer, a programming language and
a compiler. In other words, today software is fundamentally
malleable in a way that twentieth-century industrially produced
objects were not. (The emergence of consumer 3D printing and the
open hardware movement promise to bring such exibility to
physical objects as well, but it will be a while before you can print
a whole ready-to-drive-car on your home 3D printer.)
9781623568177_txt_print.indd 93 29/04/2013 11:50
Although Turing and Von Neumann formulated this funda-
mental extendibility of software in theory, its contemporary
practicehundreds of thousands of people daily involved in
extending the capabilities of computational mediais a result of
a long historical development. This development took us from the
few early room-sized computers, which were not easy to reprogram
to a wide availability of cheap computers and programming tools
decades later. This democratization of software development was
at the core of Kays vision. Kay was particularly concerned with
how to structure programming tools in such a way that would
make development of media software possible for ordinary users.
For instance, at the end of the 1977 article I have already exten-
sively quoted, he and Goldberg write, We must also provide
enough already-written general tools so that a user need not start
from scratch for most things she or he may wish to do.
Comparing the process of continuous media innovation via new
software to the history of earlier, pre-computational media reveals a
new logic at work. According to a commonplace idea, when a new
medium is invented it rst closely imitates already existing media,
before discovering its own language and aesthetics. Indeed, the
rst Gutenberg Bible closely imitated the look of the handwritten
manuscripts; early lms produced in the 1890s and 1900s mimicked
the presentational format of theatre by positioning the actors on the
invisible shallow stage and having them face the audience. Slowly,
printed books developed a diferent way of presenting information;
similarly cinema also developed its own original concept of narrative
space. Through repetitive shifts in points of view presented in subse-
quent shots, the viewers were placed inside this spacethus literally
nding themselves inside the story.
Can this logic apply to the history of computer media? As
theorized by Turing and Von Neumann, the computer is a general-
purpose simulation machine. This is its uniqueness and its diference
from all other machines and previous media. This means that the
idea that a new medium gradually nds its own language cannot
apply to computer media. If this were true it would go against
the very denition of a modern digital computer. This theoretical
argument is supported by practice. The history of computer media
so far has been not about arriving at some standardized language
as, for instance, happened with cinemabut rather about the
gradual expansion of uses, techniques, and possibilities. Rather
9781623568177_txt_print.indd 94 29/04/2013 11:50
than arriving at a particular language, we are gradually discovering
that the computer can speak more and more languages.
If we are to look more closely at the early history of computer
mediafor instance, the way we have been looking at Kays ideas
and work in this textwe will discover another reason why the idea
of a new medium gradually discovering its own language does not
apply to computer media. The systematic practical work on making
a computer simulate and extend existing media (Sutherlands
Sketchpad, the rst interactive word processor developed by
Engelbarts group, etc.) came after computers had already been
put to multiple usesperforming diferent types of calculations,
solving mathematical problems, controlling other machines in
real time, running mathematical simulations, simulating some
aspects of human intelligence, and so on. (We should also mention
the work on SAGE by MIT Lincoln Laboratory which, by the
middle of the 1950s, had already established the idea of inter-
active communication between a human and a computer via a
screen with a graphical display and a pointing device. In fact,
Sutherland developed Sketchpad on a TX-2, the new version of
a larger computer MIT constructed for SAGE.) Therefore, when
the generation of Sutherland, Nelson, and Kay started to create
new media, they built it on top, so to speak, of what computers
were already known to be capable of. Consequently they added
new properties into physical media they were simulating right
away. This can be very clearly seen in the case of Sketchpad.
Understanding that one of the roles a computer can play is that
of a problem solver, Sutherland built in a powerful new feature
that never before existed in a graphical mediumsatisfaction of
constraints. To rephrase this example in more general terms, we
can say that rather than moving from an imitation of older media
to nding its own language, computational media was from the
very beginning speaking a new language.
In other words, the pioneers of computational media did
not have the goal of making the computer into a remediation
machine which would simply represent older media in new
ways. Instead, knowing well the new capabilities provided by
digital computers, they set out to create fundamentally new kinds
of media for expression and communication. These new media
would use as their raw content the older media which already
served humans well for hundreds and thousands of yearswritten
9781623568177_txt_print.indd 95 29/04/2013 11:50
language, sound, line drawings and design plans, and continuous
tone images (i.e. paintings and photographs). But this does not
compromise the newness of new media. Computational media uses
these traditional human media simply as building blocks to create
previously unimaginable representational and information struc-
tures, creative and thinking tools, and communication options.
Although Sutherland, Engelbart, Nelson, Kay, and others
developed computational media on top of already existing devel-
opments in computational theory, programming languages, and
computer engineering, it would be incorrect to conceive the history
of such inuences as only going in one directionfrom already
existing and more general computing principles to particular
techniques of computational media. The inventors of computa-
tional media had to question many, if not most, already established
ideas about computing. They have dened many new fundamental
concepts and techniques of how both software and hardware
function, thus making important contributions to hardware and
software engineering. A good example is Kays development of
Smalltalk, which for the rst time systematically established a
paradigm of object-oriented programming. Kays rationale to
develop this new programming language was to give a unied
appearance to all applications and the interface of the PARC
system and, even more importantly, to enable its users to quickly
program their own media tools. (According to Kay, an object-
oriented illustration program written in Smalltalk by a particularly
talented 12-year-old girl was only a page long.
) Subsequently
the object-oriented programming paradigm became very popular
and object-oriented features have been added to most popular
languages such as C.
Looking at the history of computer media and examining the
thinking of its inventors makes it clear that we are dealing with the
opposite of technological determinism. When Sutherland designed
Sketchpad, Nelson conceived hypertext, Kay programmed a paint
program, and so on, each new property of computer media had
to be imagined, implemented, tested, and rened. In other words,
these characteristics did not simply come as an inevitable result
of a meeting between digital computers and modern media.
Alan Kay, Doing with Images Makes Symbols (University Video Communications,
1987), videotaped lecture, http://archive.org/details/AlanKeyD1987/
9781623568177_txt_print.indd 96 29/04/2013 11:50
Computational media had to be invented, step-by-step. And it was
invented by people who were looking for inspiration in modern
art, literature, cognitive and education psychology, and theory
of media as much as technology. For example, Kay recalls that
reading McLuhans Understanding Media led him to a realization
that a computer can be a medium rather than only a tool.

Accordingly, the opening section of Kay and Goldbergs article is
called Humans and Media, and it does read like media theory.
But this is not a typical theory that only describes the word, as
it currently exists. Similar to Marxs analysis of capitalism in his
works, here the analysis is used to create a plan for action for
building a new worldin this case, enabling people to create new
But the most important example of such non-deterministic
development is the invention of the modern interactive graphical
human-computer interface itself by Sutherland, Engelbart, Kay
and others. None of the key theoretical concepts of modern
computing as developed by Turing and Von Neumann called for
an interactive interface. In the late 1940s and 1950s the MIT
Lincoln Laboratory developed interactive graphical computers
used in SAGEthe control centers created around the US to
collect information from radar stations and coordinate a counter-
attack. But the SAGE interface was designed for very particular
tasks and it had no efect on the development of commercial
computing. It did, however, lead to a new smaller machine: the
TX-2, used by young students at MIT (including Sutherland) to
explore what can be done with an interactive computeri.e.
a computer which had a visual display. Some students started to
create interactive games including the famous Spacewar (1960).
Sutherland was one of these students who were exploring the
possibilities of visual interactive computing using the TX-2. He
went to create Sketchpad (his Ph.D. thesis) which inuenced other
pioneers of cultural computing in the 1960s including Kay. But
the theoretical road that led from SAGE to modern GUI through
PARC was a very long one.
According to Kay, the key step for him and his group was
to start thinking about computers as a medium for learning,
Alan Kay, User Interface: A Personal View, p. 1923.
9781623568177_txt_print.indd 97 29/04/2013 11:50
experimentation, and artistic expression which can be used not
just by adults but also by children of all ages.
Kay was strongly
inuenced by the theory of the cognitive psychologist Jerome
Bruner. Bruner developed his theory by redening the ideas of
Jean Piaget who postulated that children go through a number of
distinctive intellectual stages as they develop: a kinesthetic stage,
a visual stage, and a symbolic stage. But while Piaget thought
that each stage only exists for a particular period during a childs
development only to be completely replaced by a new stage, Bruner
suggested that separate mentalities that correspond to these stages
continue to exist as the child grows. That is, the mentalities do not
replace each other but are added. Bruner gave slightly diferent
names to these diferent mentalities: enactive, iconic, and symbolic.
While each mentality has developed at diferent stages of human
evolution, they continue to co-exist in an adult.
Kays interpretation of this theory was that a user interface
should appeal to all these three mentalities. In contrast to a
command-line interface, which is not accessible for children and
forces the adult to use only symbolic mentality, the new interface
should also make use of emotive and iconic mentalities. Kay also
drew on a number of studies on creativity in math, science, music,
art and other areas which suggested that initial creative work
is done mostly in iconic mentality and also in enactive.
provided additional motivation for the idea that if computers were
to function as a dynamic medium for learning and creativity they
should allow their users to think not only through symbols but also
through actions and images.
Following Kays interpretation of Bruners work, the group
at PARC mapped Bruners theory of multiple mentalities into
the interface technologies in the following way.

Mouse activates
enactive mentality (know where you are, manipulate). Icons and
windows activate iconic mentality (recognize, compare, congure.)
Finally, Smalltalk programming language allows for the use
Alan Kay, A Personal Computer for Children of All Ages, Proceedings of
the ACM National Conference, Boston, 1972, http://www.mprove.de/diplom/gui/
Alan Kay, User Interface: A Personal View, p. 195.
9781623568177_txt_print.indd 98 29/04/2013 11:50
of symbolic mentality (tie together long chains of reasoning,

In actual use, a contemporary GUI involves constant interplay
between diferent mentalities. You use a mouse to move around
the screen as though it is a physical space and point at screen
objects. All objects are represented by visual icons. You double-
click on an icon to activate it or, if it is a folder icon, to examine
its contents. This can be interpreted as an equivalent of picking
up and examining a physical object in a real world. After a folder
window opens, you may switch between diferent views, looking
at the data as icons and alternatively as a list, then sort the list
in diferent ways to examine le names, creation dates and other
symbolic information (i.e. text). If you did not nd the les you
were looking for, you may then use a search function to search the
whole computerpossibly dening multiple options and carefully
choosing the search terms (symbolic mentality). As these examples
demonstrate, the user is constantly switching between diferent
mentalities using whatever works best at a given moment.
But in addition to the general interface principles, other key
techniques that were developed by Kays group can also be under-
stood as enabling the use of diferent mentalities in combination
with each other. For instance, the user interface developed at PARC
was the rst to run on a bit-mapped displaywhich meant not
only giving users the ability to move the pointer and open multiple
windows but also to write simulation programs in Smalltalk which
could display their results visually right on the screen. By making
a change in the code a user would be able to see the visual result
of this change in the image produced by the program. Today this
ability is fundamental to computer use in all areas of science (in
particular, the use of interactive visualization and data analysis
software). And of course, we should not forget about all the media
editors created at PARC: a paint program, an illustration program,
a music editor, etc. These media editors gave the users the ability
to switch between diferent mentalities in a way not available in
the physical media. For instance, the objects in the animation
program could be drawn by hand or by writing code in Smalltalk.
As Kay and Goldberg point out, The control of the animation
Ibid., p. 197.
9781623568177_txt_print.indd 99 29/04/2013 11:50
could be easily done from a Smalltalk simulation. For example,
an animation of objects bouncing in a room is most easily accom-
plished by a few lines of Smalltalk code that express the class of
bouncing objects in physical terms.
In dening this new type of user interface, Kay and his collabo-
rators simultaneously created a radically new type of media. If we
are to agree with Bruners theory of multiple mentalities and Kays
interpretation of this theory, we should conclude that the new
computational media that he helped to invent can do something
no previous media canactivate our multiple mentalities which
all play a role in learning and creativity, allowing a user to
employ whatever works best at any given moment and to rapidly
switch between them as necessary. This may explain the success
and popularity of the GUI, which, forty years after its invention,
continues to dominate our interaction with computers. People
prefer it not because it is easy or seamless or intuitive. It
is successful because it was designed to help them think, discover,
and create new concepts using not just one type of mentality but all
of them together. In short, while many HCI experts and designers
continue to believe that the ideal human-computer interface should
be invisible and get out of the way to let users do their work,
looking at the theories of Kay and Goldberg that were behind GUI
design gives a very diferent way of understanding an interfaces
identity. Kay and his colleagues at PARC have conceived GUI as a
medium designed in its every detail to facilitate learning, discovery,
and creativity.
Given the overall emphasis of information society on constant
innovation, continuous learning, and creativity, it is only appro-
priate that as this society was coming into existence, a new
medium was being invented specically to facilitate these needs.
In 1973 Daniel Bell published his highly inuential The Coming
of Post-Industrial Society; right around that time at PARC Kay,
Goldberg, Chuck Thacker, Dan Ingalls, Larry Tesler, and other
members of the Learning Research Group created the paradigm
of modern computing. Or rather, they reinvented the computer
from a fast calculator that can only work on tasks articulated
Kay and Goldberg, Personal Dynamic Media, p. 399.
9781623568177_txt_print.indd 100 29/04/2013 11:50
beforehand to an interactive support system for thinking and
discovery. In short: from a tool to a metamedium.
Unfortunately, when GUI became the commercially successful
paradigm following the success of Apples Mac computers, intro-
duced in 1984, the intellectual origins of GUI were forgotten.
Instead, GUI was justied using a simplistic idea that since
computers are unfamiliar to people, we should help them by
making interface intuitive by making it mimic something users
are already well familiar withthe physical world outside of a
computer (which in reality was an ofce environment with folders,
desks, printers, etc.) Surprisingly, even in recent years when
born digital generations were already using computer devices
even before they ever set foot in an ofcethis idea was still used
to explain GUI. For example, Apples iPhone Human Interface
guidelines (March 2010) advise developers: When possible, model
your applications objects and actions on objects and actions in the
real world. This technique especially helps novice users quickly
grasp how your application works. Folders are a classic software
metaphor. People le things in folders in the real world, so they
immediately understand the idea of putting data into folders on
a computer.
The irony of this statement is that these Interface
guidelines are also aimed at the developers of iPadwhich clearly
represents yet another step in migration from the world of physical
print to all-digital environment. It is as though we are asked to
remember and cherish the older mediaand erase it at the same
The computer as a metamedium
As we have established, the development of computational media
runs contrary to previous media history. But in a certain sense,
the idea of a new media gradually discovering its own language
actually does apply to the history of computational media after
all. And just as with printed books and cinema, this process took a
html#//apple_ref/doc/uid/TP40006556-CH7-SW1 (April 5, 2010).
9781623568177_txt_print.indd 101 29/04/2013 11:50
few decades. When the rst computers were built in the middle of
the 1940s, they could not be used as media for cultural represen-
tation, expression, and communication. Slowly, through the work
of Sutherland, Engelbart, Nelson, Papert, and others in the 1960s,
the ideas and techniques were developed that made computers
into a cultural machine. One could create and edit text, make
drawings, move around a virtual object, etc. And nally, when
Kay and his colleagues at PARC systematized and rened these
techniques and put them under the umbrella of a GUI (making
computers accessible to multitudes) a digital computer nally was
given its own languagein cultural terms. In short, only when
a computer became a cultural mediumrather than merely a
versatile machinecould it be so used.
Or rather, it became something that no other media had been
before. For what had emerged was not yet another media, but as
Kay and Goldberg insist in their article, something qualitatively
diferent and historically unprecedented. To mark this diference,
they introduce a new termmetamedium.
This metamedium is unique in a number of diferent ways. One of
them I have already discussed in detailit can represent most other
media while augmenting them with many new properties. Kay and
Goldberg also name other properties that are equally crucial. The
new metamedium is activeit can respond to queries and experi-
mentsso that the messages may involve the learner in a two-way
conversation. For Kay who was strongly interested in children
and learning, this property was particularly important since, as
he puts it, it has never been available before except through the
medium of an individual teacher.
Further, the new metamedium
can handle virtually all of its owners information-related needs.
(I have already discussed the consequence of this property above.)
It can also serve as a programming and problem solving tool
and an interactive memory for the storage and manipulation of
But the property that is the most important from the point
of view of media history is that the computer metamedium is simul-
taneously a set of diferent media and a system for generating new
media tools and new types of media. In other words, a computer
Kay and Goldberg, Personal Dynamic Media, p. 394.
Ibid., p. 393.
9781623568177_txt_print.indd 102 29/04/2013 11:50
can be used to create new tools for working with the media types it
already provides as well as to develop new not-yet-invented media.
In the opening to his book Expressive Processing, Noah Wardrip-
Fruin perfectly articulates this meta-generative specicity of
A computer can simulate a typewritergetting input from the
keyboard and arranging pixels on the screen to shape the corre-
sponding lettersbut it can also go far beyond a typewriter,
ofering many fonts, automatic spelling correction, painless
movement of manuscript sections (through simulations of cut
and paste), programmable transformations (such as nd and
replace), and even collaborative authoring by large, dispersed
groups (as with projects like Wikipedia). This is what modern
computers (more lengthily called stored-program electronic
digital computers) are designed to make possible: the continual
creation of new machines, opening new possibilities, through
the denition of new sets of computational processes.
Using the analogy with print literacy, Kay motivates this property
in this way: The ability to read a medium means you can access
materials and tools generated by others. The ability to write in a
medium means you can generate materials and tools for others.
You must have both to be literate.
Accordingly, Kays key efort
at PARC was the development of the Smalltalk programming
language. All media editing applications and the GUI itself were
written in Smalltalk. This made all the interfaces of all applica-
tions consistent, facilitating quick learning of new programs.
Even more importantly, according to Kays vision, Smalltalk
would allow even novice users to write their own tools and dene
their own media. In other words, all media editing applications
that would be provided with a computer, were to serve also as
examples, inspiring users to modify them and to write their own
Noah Wardrip-Fruin, Expressive Processing: Digital Fictions, Computer Games,
and Software Studies (The MIT Press, 2009).
Alan Kay, User Interface: A Personal View, in The Art of Human-Computer
Interface Design, ed. Brenda Laurel (Reading, MA, Addison-Wesley, 1990), p. 193.
The emphasis is in the original.
9781623568177_txt_print.indd 103 29/04/2013 11:50
Accordingly, the large part of Kay and Goldbergs paper is
devoted to description of software developed by the users of
their system: an animation system programmed by animators,
a drawing and painting system programmed by a child, a
hospital simulation programmed by a decision-theorist, an audio
animation system programmed by musicians, a musical score
capture system programmed by a musician, electronic circuit
design by a high school student. As can be seen from this list,
(which corresponds to the sequence of examples in the article), Kay
and Goldberg deliberately juxtaposed diferent types of users
professionals, high school students, and childrenin order to
show that everybody could develop new tools using the Smalltalk
programming environment.
The sequence of examples also strategically juxtaposes media
simulations with other kinds of simulations in order to emphasize
that simulation of media is only a particular case of the computers
general ability to simulate all kinds of processes and systems. This
juxtaposition of examples gives us an interesting way to think
about computational media. Just as a scientist may use simulation
to test diferent conditions and play diferent what/if scenarios, a
designer, a writer, a musician, a lmmaker, or an architect working
with computer media can quickly test diferent creative direc-
tions in which the project can be developed as well as see how
modications of various parameters afect the project. The latter
is particularly easy today since the interfaces of most media editing
software not only explicitly present these parameters but also
simultaneously give the user the controls for their modication. For
instance, when the Formatting Palette in Microsoft Word shows
the font used by the currently selected text, it is displayed in a
column next to all other fonts available. Trying diferent fonts is as
easy as scrolling down and selecting the name of a new font.
Giving users the ability to write their own programs was a crucial
part of Kays vision for the new metamedium he was inventing
at PARC. According to Noah Wardrip-Fruin, Engelbarts research
program was focused on a similar goal: Engelbart envisioned users
creating tools, sharing tools, and altering the tools of others.

Noah Wardrip-Fruin, introduction to Douglas Engelbart and William English,
A Research Center for Augmenting Human Intellect (1968), New Media Reader,
p. 232.
9781623568177_txt_print.indd 104 29/04/2013 11:50
Unfortunately, when in 1984 Apple shipped Macintosh, which
was to become the rst commercially successful personal computer
modeled after the PARC system, it did not have an easy-to-use
programming environment. HyperCard, written for Macintosh in
1987 by Bill Atkinson (who was one of PARCs alumni), gave users
the ability to quickly create certain kinds of applicationsbut it did
not have the versatility and breadth envisioned by Kay. Only more
recently, as the general computer literacy has widened and many
new high-level programming languages have become available
Perl, PHP, Python, JavaScript, etc.have more people started to
create their own tools by writing software. A good example of a
contemporary programming environment, very popular among
artists and designers and which, in my view, is close to Kays vision,
is Processing.
Built on top of the Java programming language,
Processing features a simplied programming style and an extensive
library of graphical and media functions. It can be used to develop
complex programs and also to quickly test ideas. Appropriately, the
ofcial name for Processing projects is sketches.
In the words of
Processing inventors and main developers Ben Fry and Casey Reas,
the languages focus is on the process of creation rather than
end results.
Another popular programming environment that
similarly enables quick development of media projects is Max/MSP
and its successor PDboth developed by Miller Puckette.
At the end of the 1977 article that served as the basis for our
discussion in this chapter, Kay and Goldberg summarize their
arguments in the phrasewhich in my view is the best formulation
we have had so farof what computational media is artistically
and culturally. They call the computer a metamedium whose
content is a wide range of already-existing and not-yet-invented
media. In another article published in 1984 Kay unfolds this
denition. As a way of concluding this chapter, I would like to
quote this longer denition which is as accurate and inspiring
today as it was when Kay wrote it:
It [a computer] is a medium that can dynamically simulate the
details of any other medium, including media that cannot exist
9781623568177_txt_print.indd 105 29/04/2013 11:50
physically. It is not a tool, though it can act like many tools. It
is the rst metamedium, and as such it has degrees of freedom
for representation and expression never before encountered and
as yet barely investigated.
Alan Kay, Computer Software, Scientic American (September 1984), p. 52.
Quoted in Jean-Louis Gasse, The Evolution of Thinking Tools, in The Art of
Human-Computer Interface Design, p. 225.
9781623568177_txt_print.indd 106 29/04/2013 11:50
Understanding metamedia
It [the electronic book] need not be treated as a simulated
paper book since this is a new medium with new properties.
Kay and Goldberg, Personal Dynamic Media, 1977
Today Popular Science, published by Bonnier and the largest
science+tech magazine in the world, is launching Popular
Science+ the rst magazine on the Mag+ platform, and you
can get it on the iPad tomorrowWhat amazes me is that you
dont feel like youre using a website, or even that youre using
an e-reader on a new tablet device which, technically, is what
it is. It feels like youre reading a magazine. (emphasis is in the
Popular Science+, posted on April 2, 2010.
The building blocks
I started putting this book together in 2007. Today is April 3, 2010,
and I am editing this chapter. Today is also an important day in the
history of media computing (which started exactly forty years ago
with Ivan Sutherlands Sketchpad)Apples iPad tablet computer
rst went on sale in the US on this date. During the years I was
writing and editing the book, many important developments made
Alan Kays vision of a computer as the rst metamedium more
realand at the same time more distant.
9781623568177_txt_print.indd 107 29/04/2013 11:50
The dramatic cuts in the prices of laptops and the rise of cheap
notebooks (and, in the years that followed, tablet computers),
together with the continuing increase in the capacity and decrease
in price of consumer electronics devices (digital cameras, video
cameras, media players, monitors, storage, etc.) brought media
computing to even more people. With the price of a notebook ten or
twenty times less than the price of a large digital TV set, the 1990s
argument about the digital divide became less relevant. It became
cheaper to create your own media than to consume professional
TV programs via the industrys preferred mode of distribution.
More students, designers, and artists learned Processing and
other specialized programming and scripting languages speci-
cally designed for their needswhich made software-driven art
and media design more common. Perhaps most importantly, most
mobile phones became smart phones supporting Internet connec-
tivity, web browsing, email, photo and video capture, and a range
of other media creation capabilitiesas well as the new platforms
for software development. For example, Apples iPhone went on
sale on June 29, 2007; on July 10 when the App Store opened,
it already had 500 third-party applications. According to Apples
statistics, on March 20, 2010 the store had over 150,000 diferent
applications and the total number of application downloads had
reached 3 billion. In February 2012, the numbers of iOS apps
reached 500,000 (not counting many more that Apple rejected),
and the total number of downloads was already 25 billion.
At the same time, some of the same developments strengthened
a diferent vision of media computinga computer as a device for
buying and consuming professional media, organizing personal
media assets and using GUI applications for media creation and
editingbut not imagining and creating not-yet-invented media.
Apples rst Mac computer, released in 1984, did not support
writing new programs to take advantage of its media capacities.
The adoption of the GUI interface for all PC applications by the
software industry made computers much easier to use but in the
same time took away any reason to learn programming. Around
2000, Apples new paradigm of a computer as a media hub (or
a media center)a platform for managing all personally created
http://www.apple.com/itunes/25-billion-app-countdown/ (March 5, 2012).
9781623568177_txt_print.indd 108 29/04/2013 11:50
mediafurther erased the computer part of a PC. During the
following decade, the gradual emergence of web-based distribution
channels for commercial media, such as Apple iTunes Music Store
(2003), internet television (in the US the rst successful service was
Hulu, publically launched on March 12, 2008), the e-book market
(Random House and HarperCollins started selling their titles in
digital form in 2002) and nally the Apple iBookstore (April 3,
2010), together with specialized media readers and players such
as Amazons Kindle (November 2007) have added a new crucial
part to this paradigm. (In the early 2010s we also got the Android
app market, the Amazon app store, etc.) A computer became even
more of a universal media machine than beforewith the focus
on consuming media created by others.
Thus, if in 1984 Apples rst Apple computer was critiqued for
its GUI applications and lack of programming tools for the users,
in 2010 Apples iPad was critiqued for not including enough GUI
tools for heavy duty media creation and editinganother step
back from Kays Dynabook vision. The following quote from an
iPad review by Walter S. Mossberg from the Wall Street Journal
was typical of journalists reactions to the new device: if youre
mainly a Web surfer, note-taker, social-networker and emailer, and
a consumer of photos, videos, books, periodicals and musicthis
could be for you.
The New York Times NYTs David Pogue
echoed this: The iPad is not a laptop. Its not nearly as good for
creating stuf. On the other hand, its innitely more convenient
for consuming itbooks, music, video, photos, Web, e-mail and
so on.
Regardless of how much contemporary universal media
machines fulll or betray Alan Kays original vision, they are
only possible because of it. Kay and others working at Xerox
PARC built the rst such machine by creating a number of media
authoring and editing applications with a unied interface, as
well as the technology to enable the machines users to extend its
capacities. Starting with the concept Kay and Goldberg proposed
in 1977, to sum up this work at PARC (the computer as a
metamedium whose content is a wide range of already-existing
and not-yet-invented media) in this chapter I will discuss how this
http://ptech.allthingsd.com/20100331/apple-ipad-review/ (April 3, 2010).
http://gizmodo.com/5506824/first-ipad-reviews-are-in (April 3, 2010).
9781623568177_txt_print.indd 109 29/04/2013 11:50
concept redenes what media is. In other words, I will go deeper
into the key question of this book: what exactly is media after
Approached from the point of view of media history, the
computer metamedium contains two diferent types of media.
The rst type is simulations of prior physical media extended
with new properties, such as electronic paper. The second type
is a number of new computational media that have no physical
precedents. Here are the examples of these new media proper,
listed with names of the people and/or places usually credited as
their inventors: hypertext and hypermedia (Ted Nelson); interactive
navigable 3D spaces (Ivan Sutherland), interactive multimedia
(Architecture Machine Groups Aspen Movie Map).
This taxonomy is consistent with the denition of the computer
metamedium given in the end of Kay and Goldbergs article. But
let us now ask a new question: what are the building blocks of
these simulations of previously existing media and newly invented
media? Actually we have already encountered these blocks in the
preceding discussion but until now I have not explicitly pointed
them out.
The building blocks used to make up the computer metamedium
are diferent types of media data and the techniques for generating,
modifying, and viewing this data. Currently, the most widely used
data types are text, vector images and image sequences (vector
animation), continuous tone images and sequences of such images
(i.e., photographs and digital video), 3D models, geo-spatial data,
and audio. I am sure that some readers would prefer a somewhat
diferent list and I will not argue with them. What is important at
this point for our discussion is to establish that we have multiple
kinds of data rather than just one kind.
This points leads us to the next one: the techniques for data
manipulation themselves can be divided into two types depending
on which data types they can work on:
(A) The rst type is media creation, manipulation, and access
techniques that are specic to particular types of data.
In other words, these techniques can be used only on
a particular data type (or a particular kind of media
content). I am going to refer to these techniques as media-
specic (the word media in this case really stands for
9781623568177_txt_print.indd 110 29/04/2013 11:50
data type). For example, the technique of geometrical
constraint satisfaction invented by Sutherland can work
on graphical data dened by points and lines. However,
it would be meaningless to apply this technique to text.
Another example: today image editing programs usually
include various lters such as blur and sharpen which
can operate on continuous tone images. But normally
we would not be able to blur or sharpen a 3D model.
Similarly, it would be as meaningless to try to extrude a
text or interpolate it as to dene a number of columns
for an image or a sound composition.
Some of these data manipulation techniques appear
to have no historical precedents in physical mediathe
technique of geometric constraint satisfaction is a case
in point. Another example of such new technique is
evolutionary algorithms commonly used to generate still
images, animations, and 3D forms. Other media-specic
techniques do refer to prior physical tools or machinesfor
instance, brushes in image editing applications, a zoom
command in graphics software, or a trim command in
video editing software. In other words, the same division
between simulations and properly new media also
applies to the individual techniques that make up the
computer metamedium.
(B) The second type is new software techniques that can
work with digital data in general (i.e. they are not media-
specic). The examples are view control, hyperlinking,
sort, search, network protocols such as HTTP, and various
data analysis techniques from the elds of Articial
Intelligence, Machine Leaning, Knowledge Discovery, and
other sub-elds of computer science. (In fact, large parts
of computer science, information science and computer
engineering science are about these techniquessince they
focus on designing algorithms for processing information in
general.) These techniques are general ways of manipulating
data regardless of what this data encodes (i.e. pixel values,
text characters, sounds, etc.). I will be referring to these
techniques as media-independent. For instance, as we
saw, Engelbarts view controlthe idea that the same
9781623568177_txt_print.indd 111 29/04/2013 11:50
information can be displayed in many diferent waysis
now implemented in most media editors and therefore
works with images, 3D models, video les, animation
projects, graphic designs, and sound compositions. View
control has also become part of the modern OS (operating
systems such as Mac OS X, Microsoft Windows, or Google
Chrome OS). We use view control daily when we change
the les view between icons, list, and columns
(these are names used in Mac OS X; other OS may use
diferent names to refer to the same views). General media-
independent techniques also include interface commands
such as cut, copy, and paste. For instance, you can select a
le name in a directory, a group of pixels in an image, or
a set of polygons in a 3D model, and then cut, copy, and
paste these selected objects.
OK: we now have two diferent ways of dividing up the
computer metamedium. If we want to continue using the concept
of a medium, we will say that a computer simulates prior
mediums and allows for the denition of new ones. Alternatively,
we can think of the computer metamedium as a collection of data
types, media-specic techniques that can only operate on particular
types, and media-independent techniques that can work on any
data. Each of the mediums contained in the computer metamedium
is made from some of these building blocks. For example, the
elements of the navigable 3D space medium are 3D models plus
techniques for representing them in perspective, texture mapping,
simulating efects of various types of lights on their surface, casting
shadows, and so on. In another example, the elements of digital
photography are continuous tone images captured by lens-based
sensors plus a variety of techniques for manipulating these images:
changing contrast and saturation, scaling, compositing, etc.
A note about the distinction between media-specic and media-
independent techniques: it works in theory. In practice, however,
it is often hard to say in what category a particular technique or
a medium should be placed. For instance, is Sketchpads ability
to work at any of 2000 levels of magnication an extension of
techniques which existed previously (moving ones body closer
to the drawing board, using a magnifying lens) or is it something
really new? Or what about 3D navigable space, which I have
9781623568177_txt_print.indd 112 29/04/2013 11:50
used as an example of a new medium only made possible by
computers (tracing it to Sutherlands rst Virtual Reality system
of 1966)? Is it newor is it an extension of a physical medium
of architecture, which allows a human being to walk around the
built structures?
The boundaries between simulated media and new media,
or between media-specic and media-independent techniques
should not be thought of as solid walls. Rather than thinking of
them as rigidly dened categories, let us imagine them as coordi-
nates of the map of the computer metamedium. Like any rst
sketch, no matter how imprecise, this map is useful because now
we have something to modify as we go forward.
Media-independent vs.
media-specic techniques
Having drawn our rst map of the computer metamedium, let us
now examine it to see if it can reveal something which we did not
notice so far. We see a number of mediums, old and newand this
certainly ts with our common understanding of media history.
(For example, take a look at the table of contents of McLuhans
Understanding Media and you will nd two dozen chapters each
devoted to a particular mediumwhich for McLuhan range from
writing and roads to cars and TV.) We also see various media-
specic techniques and this again is something we are familiar
with: think of editing techniques in cinema, dening a contour in
painting, creating rhyme in poetry, or shaping a narrative out of
chronological story events in literature. But one area of the map
does looks new and diferent in relation to previous cultural history.
This is the area that contains media-independent techniques.
What are these techniques, and how can they work across media,
i.e. on diferent types of data? (Design across media was a phrase
used by Adobe in marketing an early version of its Creative Suite
of media authoring applications.)
I am going to argue that media independence does not just
happen by itself. For a technique to work with various data types,
programmers have to implement a diferent method for each data
type. Thus, media-independent techniques are general concepts
9781623568177_txt_print.indd 113 29/04/2013 11:50
translated into algorithms, which can operate on particular data
types. Let us look at some examples.
Consider the omnipresent cut and paste. The algorithm to select
a word in a text document is diferent from the algorithm to select
a curve in a vector drawing, or the algorithm to select a part of a
continuous tone (i.e. raster) image. In other words, cut and paste
is a general concept that is implemented diferently in diferent
media software depending on which data type this software is
designed to handle. (In Larry Teslers original implementation of
the universal commands concept done at PARC in 19745, it only
worked for text editing.) Although cut, copy, paste, and a number
of similar universal commands are available in all contemporary
GUI applications for desktop computers (but not necessarily in
mobile phone apps), what they actually do and how they do it is
diferent from application to application.
Search operates in the same way. The algorithm to search for a
particular phrase in a text document is diferent than the algorithm
that searches for a particular face in a photo or a video clip. (I am
talking here about content-based search, i.e. the type of search
which looks for information inside actual images, as opposed to
only searching image titles and other metadata the way image
search engines such as Google Image Search were doing it in the
2000s.) However, despite these diferences the general concept
of search is the same: locating any elements of a single media
objector any media objects in a larger setto match particular
user-dened criteria. Thus we can ask the web browser to locate
all instances of a particular word in a current web page; we can
ask a web search engine to locate all web pages which contain a set
of keywords; and we can ask a content-based image search engine
to nd all images that are similar in composition to an image we
Because of the popularity of the search paradigm on the web,
we now assume that in principle we canor will be able to in the
futuresearch any media. In reality it is much easier to search data
that has a modular organizationsuch as text or 3D modelsthan
media that does not have it, such as continuous-tone images, video,
or audio. But for the users these diferences are not importantas
far as they are concerned, all types of media content acquire a new
common property that can be called searchability.
Similarly, in the mid-2000s photo and video media started
9781623568177_txt_print.indd 114 29/04/2013 11:50
to acquire another property of ndability. (I am borrowing this
term from 2005 book by Peter Morville, Ambient Findability:
What We Find Changes Who We Become
). The appearance of
consumer GPS-enabled media capture devices and the addition of
geo-tagging, geo-search, and mapping services to media sharing
sites such as Flickr (added in 2006) and media management appli-
cations such as iPhoto (added in 2009) gradually made media
location aware.
Another example of a general concept that, through the eforts
by many people, was gradually made to work with diferent media
typesand thus became a media-independent technique
is information visualization (often abbreviated as infovis.) The
name infovis already suggests that it is not a media-specic
techniquerather it is a very general method that potentially can
be applied to any data. The name implies that we can potentially
take anythingnumbers, text, network, sound, video, etc.and
map it into image to reveal patterns and relationships in the data.
(A parallel to information visualization is data sonication, which
renders data as sound).
However, it took decades to invent techniques to turn this
potential into reality. In the 1980s the emerging eld of scientic
visualization focused on 3D visualization of numerical data. In
the second part of the 1990s the growing graphics capabilities of
PCs made possible for larger numbers of people to experiment
with visualizationwhich led to the development of techniques to
visualize media. The rst successful visualizations of large bodies of
text appeared around 1998 (Rethinking the Book by David Small,
1998; Valence by Ben Fry, 1999; TextArc by W. Bradford Paley,
); visualizations of musical structures in 2001 (The Shape of
Song by Martin Wattenberg); and visualization of a feature lm in
2000 (The Top Grossing Film of All Time, 1 x 1 by Jason Salovan).
Information visualization is a particularly interesting example
of a new media-independent technique because of the variety
of the algorithms and strategies for representing data visually.
Peter Morville. Ambient Findability: What We Find Changes Who We Become.
OReilly Media, Inc., 2005.
W. Bradford Paley, TextArc, 2002, http://www.textarc.org/; Ben Fry, Valence,
1999, http://benfry.com/valence/; David Small, Rethinking the Book, PhD thesis,
1999, http://acg.media.mit.edu/projects/thesis/DSThesis.pdf
9781623568177_txt_print.indd 115 29/04/2013 11:50
Wind Map, a real-time dynamic visualization of the wind currents over
the USA. Fernanda Vigas and Martin Wattenberg, 2012.
9781623568177_txt_print.indd 116 29/04/2013 11:50
9781623568177_txt_print.indd 117 29/04/2013 11:50
For example, Martin Wattenberg whose work, in his own words,
focuses on visual explorations of culturally signicant data,

created visualizations of a history of net art, the music compositions
of Bach, Philip Glass and other composers, the thought process of
a computer chess-playing program, and the history of Wikipedia
pages, among other projects. In each case he had to decide which
dimensions of the data to choose and how to translate them in
a visual form. But despite the diferences, we recognize all these
projects as information visualizations. They are all realizations of
the same general conceptselecting some dimensions of the data
and representing them visually through the relations of graphic
They also all rely on the same fundamental capacities
of software to manipulate numerical data and to map it from one
form to another. Finally, they all can be also understood as the
application of the new computer media of computer graphics
generation of images from numerical data. (Think of an afnity
between a 3D computer model based on a 3D scan of a face, and
a vector visualization of, for instance, the faces position over time,
based on the data extracted from a video.)
As the result of infovis work by Wattenberg and other people
over the course of last two decades many types of data acquired
a new common propertytheir structure can be visualized. This
new property of media is distributed across various applications,
software libraries, art and design projects, research papers, and
prototypes. Today some visualization tools are included in media
editing softwarefor instance, media editors such as Photoshop
can display a histogram of an image, Final Cut and other profes-
sional video editing software can visualize the color content of a
video clip, and many media players including iTunes ofer a music
visualization feature. Google Trends visualizes search patterns;
YouTube and Flickr visualize viewing stats for video and photos.
Going through the thousands of infovis projects collected on
infosthetics.com, visualcomplexity.com, and other blogs about
visualizations, we nd a variety of experiments in visualization of
media such as songs, poems and novels and every possible kind
http://www.bewitched.com/about.html (July 23, 2006).
For a detailed discussion of infovis, most general principles and new develop-
ments, see my article What is Visualization? (2010), Visual Studies 26, no. 1
(March 2011).
9781623568177_txt_print.indd 118 29/04/2013 11:50
of datafrom the artists sons, daughters and cats movements
in their living room over a period of an hour (1hr in front of the
TV by umblebee, 2008) to a citation network in science journals
(Eigenfactor.org by Moritz Stefaner, 2009).
We can also nd
such projects in art exhibitions such as MOMAs 2008 Design and
Elastic Mind,
SIGGRAPH 2009 Info-Aesthetics,
and MOMAs
2011 Talk to Me.
Visualization, searchability, ndabilitythese and many other
new media-independent techniques (i.e. concepts implemented
to work across many data types) clearly stand out in the map of
the computer metamedium we have drawn because they go against
our habitual understanding of media as plural (i.e. as consisting of
a number of separate mediums). If we can use the same techniques
across diferent media types, what happens to these distinctions
between mediums?
The idea that all artworks fall into a number of distinct mediums
each with its own distinct techniques and representational devices
was central to modern art and aesthetics. In his 1766 Laokoon oder
ber die Grenzen der Malerei und Poesie (Laocoon: An Essay on
the Limits of Painting and Poetry) German philosopher Gotthold
Ephraim Lessing argued for the radical diference between poetry
and painting since one is extended in time and the other is in
space. The idea reached its extreme in the rst two decades of
the twentieth century when modernist artists focused their energy
on discovering a unique language of each artistic medium. The
following statement made in 1924 by Jean Epstein, a French avant-
garde lmmaker and theoretician, is typical of modernist rhetoric
of purity; countless statements like it appeared on the pages of
avant-garde publications of the time:
For every art builds its forbidden city, its own exclusive domain,
autonomous, specic and hostile to anything that does not
belong. Astonishing to relate, literature must rst and foremost
be literary; the theater, theatrical; painting, pictorial; and the
cinema, cinematic. Painting today is freeing itself from many of
http://well-formed.eigenfactor.org/; http://www.flickr.com/photos/the_bumblebee
9781623568177_txt_print.indd 119 29/04/2013 11:50
its representational and narrative concerns The cinema must
seek to become, gradually and in the end uniquely, cinematic; to
employ, in other words, only photogenic elements.
In relation to painting, the doctrine of media purity reaches its
extreme expression in the famous argument of Clement Greenberg
that Because atness was the only condition painting shared with
no other art, Modernist painting oriented itself to atness as it
did to nothing else.
(Note that Greenberg did not advocate this
position as a justication for the abstract art of his contemporaries;
he only ofered this as a historical analysis of earlier modernism.)
Greenberg wrote: It was the stressing of the ineluctable atness
of the surface that remained, however, more fundamental than
anything else to the processes by which pictorial art criticized and
dened itself under Modernism. For atness alone was unique and
exclusive to pictorial art.
Only after the 1960s when installationa new art form based
on the idea of mixing diferent media and materialsgradually
became popular and accepted in the art world, did the obsession
with media-specicity lose its importance.
However, even during its dominance the principle of media-
specicity was always counterbalanced by its opposite. Throughout
the modern period we also nd locali.e. specic to particular
historical moments and artistic schoolsattempts to formulate
aesthetic principles that can relate diferent mediums to each
other. Consider for instance the considerable eforts spent by many
modernist artists to establish parallels between musical and visual
compositions. This work was often associated with the ideas of
synesthesia and Gesamtkunstwerk; it included theories, practical
compositions, and technologies such as color organs constructed
by Scriabin, the Whitneys (who went on to create the rst computer
animations) and many other artists and musicians.
While they did not explicitly theorize cross-media aesthetics
to the same degree, modernist artistic paradigmsclassicism,
Jean Epstein, On Certain Characteristics of Photognie, in French Film Theory
and Criticism, vol. 1: 190729, ed. Richard Abel (Princeton: University of Princeton
Press, 1988).
Clement Greenberg, Modern Painting, Forum Lectures (Washington, DC:
Voice of America: 1960), http://www.sharecom.ca/greenberg/modernism.html
9781623568177_txt_print.indd 120 29/04/2013 11:50
romanticism, naturalism, impressionism, socialist realism, supre-
matism, cubism, surrealism, and so oncan be also understood as
the systems which gave common properties to works in various
media. Thus, the novels of mile Zola and paintings by Manet
were aligned in their uncompromising, naturalist depiction
of ordinary people; Constructivist paintings, graphics, indus-
trial design, theatre design, architecture and fashion shared the
aesthetics of expressed structure (visually emphasizing compo-
sition structure by exaggerating it); and De Stijl aesthetics of
non-intersecting rectangular forms in primary colors were applied
to painting, furniture, architecture, and typography.
What is the diference between such earlier artistic work on
establishing media correspondences and the software techniques
that work across diferent media? Clearly, the artistic systems and
media authoring, editing and interaction techniques available in
media software operate on diferent levels. The former are respon-
sible for the content and style of the works to be generatedi.e.
what is going to be created in the rst place. The latter are used not
only to create but also to interact with what already was generated
previouslyfor instance, blogs, photos, or videos on the web
created by others.
Put diferently, the eforts by modern artists to create parallels
between mediums were prescriptive and speculative.
media properties would only apply to selected bodies of artistic
work created by particular artists or groups. In contrast, software
imposes common media properties on any media it is applied
to. Thus, software also shapes our understanding of what media
is in general. For example, web applications and services include
methods for navigating, reading, listening or viewing media
objects, attaching additional information to them (comments, tags,
geo-tagging) or nding them in a larger set (i.e. search engines and
search commands). This applies to all videos, images, text pages,
text documents, maps, etc. In other words, we can say that media
By speculative here I mean that in many cases the proposed aesthetic systems
were not fully realized in practice. For example, no purely suprematist architecture
designed by movement leader Kasimir Malevich was ever built; the same goes for
the futurist architecture of Antonio SantElia as presented in his drawings for La
Citt Nuova, 191214.
9781623568177_txt_print.indd 121 29/04/2013 11:50
software interprets any media it touches and its interpreta-
tions always include certain statements.
Of course, media-independent aesthetic systems proposed by
modernists were not only generative (the creation of new works)
but also interpretive. That is, modernist artists and theorists often
tried to change audiences understanding of past and contemporary
art, usually in a critical and negative way (each new movement
wanted to discredit its predecessors and competitors). However,
since their programs were theories rather than software, they had
no direct material efect on users interaction with the artistic
works, including those created in the past. In contrast, software
techniques afect our understanding of media through the opera-
tions they make available to us for creating, editing, interacting
with, and sharing media artifacts.
Additionally, modern artistic and aesthetic paradigms in practice
would only be realized in two, three, or maybe four mediumsbut
not all. Naturalism can be found in literature and visual arts but
not in architecture, and Constructivism did not spread to music or
literature. However cut, copy, paste and nd commands are found
in all media applications; any media object can be geo-tagged;
the view control principle is implemented to work with all media
types. To cite more examples, recall my discussion of how all media
acquire new properties such as searchability and ndability. Any
text can be searched regardless of whether it is something you
wrote or a classical novel downloaded from Project Gutenberg;
similarly, part of an image can be cut and pasted into another
image regardless of what these images are. In short: media software
afects all media content equally, regardless of its aesthetics,
semantics, authorship, and historical origin.
To summarize this discussion: in contrast to modern artistic
programs to create diferent media that share the same principles,
software media-independent techniques are ubiquitous and univer-
salist. For instance, cut and paste are built into all media editing
softwarefrom specialized professional applications to consumer
software included on every new media device sold. Further, these
techniques can be applied to any media work regardless of its
aesthetics and authorshipi.e. whether it was made by the person
who is currently applying these operations or by somebody else.
In fact, the technical ability to sample media work by others has
become the basis of the key aesthetics of our timeremixing.
9781623568177_txt_print.indd 122 29/04/2013 11:50
Of course, not all media applications and devices make all
these techniques equally availableusually for commercial and/
or copyright reasons. For example, at present the Google Books
reader does not allow you to select and copy the text from book
pages. Thus, while our analysis applies to conceptual and technical
principles of software and their cultural implications, we need to
keep in mind that in practice these principles are overwritten by
commercial structures. However, even the most restrictive software
still includes some basic media operations. By being present in
all software designed to work with diferent media types, these
operations establish a shared understanding of media today
experienced by the users as common media principles of all
To conclude this discussion of software techniques that work
across diferent media types, recall my earlier statement that
computerization of media does not collapse the diference between
mediumsbut it does bring them closer together in various ways.
(I have already provided a number of examples of this in the
beginning of my analysis of Kay and Goldbergs Personal Dynamic
Media article). Now I can name one of the key developments respon-
sible for this media attractioncommon software techniques
which can operate across diferent media types. If we recall again
Kay and Goldbergs formulation that the computer metamedium
includes a variety of already-existing and new media, this statement
can be paraphrased as follows: within the computer metamedium,
all previously existing and newly invented mediums share some
common propertiesi.e. they rely on a set of common software
techniques for data management, authoring, and communication.
It is hard to over-estimate the historical importance of the
development of these cross-media techniques. Humans have
always used some general strategies to organize their cultural
representations, experiences and actionsfor example, narrative,
symmetry, rhythm, the repetitive structures (think of ornament),
use of complementary colors, and a few others. Clearly these strat-
egies were very important for human perception, cognition, and
memoryand that is why we nd them in every culture and every
medium, from poetry to architecture, music and poetry. However,
these strategies were normally not embedded into any techno-
logical tools or materialsthey were in the minds and bodies
of artisans who were communicating them from generation to
9781623568177_txt_print.indd 123 29/04/2013 11:50
generation. Modern media for representation and communication
bring us to a new stage. They often do embody certain techniques
that apply to any media which can be generated or captured with
them (for instance, one-point linear perspective imposed by lens-
based capture technologies such as photography, lm and analog
and digital video). However these techniques would apply only
to particular media types. Against these historical developments,
the innovation of media software clearly stands. They bring a
new set of techniques which are implemented to work across all
media. Searchability, ndability, linkability, multimedia messaging
and sharing, editing, view control, zoom and other media-
independent techniques are viruses that infect everything software
touchesand therefore in their importance they can be compared
to the basic organizing principles for media and artifacts which
were used for thousands of years.
Inside Photoshop
Contemporary media is experienced, created, edited, remixed,
organized, and shared with software. This software includes stand-
alone professional media design and management applications
such as Photoshop, Illustrator, Dreamweaver, Final Cut, After
Efects, Aperture, and Maya; consumer-level apps such as iPhoto,
iMovie, or Picasa; tools for sharing, commenting, and editing
provided by social media sites such as Facebook, YouTube, Vimeo,
and Photobucket, and the numerous media viewing, editing, and
sharing apps available on mobile platforms. To understand media
today we need to understand media softwareits genealogy
(where it comes from), its anatomy (interfaces and operations),
and its practical and theoretical efects. How does media authoring
software shape the media being created, making some design
choices seem natural and easy to execute, while hiding other
design possibilities? How does media viewing/managing/remixing
software afect our experience of media and the actions we perform
on it? How does software change what media is conceptually?
This section continues investigating these general questions
that drive this book via the analysis of a software application that
became synonymous with digital mediaAdobe Photoshop.
9781623568177_txt_print.indd 124 29/04/2013 11:50
Like other professional programs for media authoring and editing,
Photoshops menus contain many dozens of separate commands. If
we consider that almost all the commands contain multiple options
that allow each command to do a number of diferent things, the
complete number runs into the thousands.
This multiplicity of operations ofered in contemporary appli-
cation software creates a challenge for Software Studies. If we are
to understand how software applications participate in shaping
our worlds and our imaginations (what people imagine they can
do with software), we need some way of sorting all these opera-
tions into fewer categories so we can start building a theory of
application software. This cannot be achieved by simply following
the top menu categories ofered by applications. (For example,
Photoshop CS4s top menu includes File, Edit, Layer, Select, Filter,
3D, View, Window, and Help.) Since most applications include
their own unique categories, our combined list will be too large. So
we need to use a more general system.
The provisional map of the computer metamedium that we
developed in previous sections provides one such possible system.
In this section we will test the usefulness of this map by analyzing a
subset of Photoshops commands which, in a certain sense, stand-in
for this application in our cultural imagination: Filters. We will also
discuss another key feature of PhotoshopLayers.
Our map organizes software techniques for working with media,
using two schemes. The rst scheme divides these techniques into
two types depending on which data types they can work on: 1)
media creation, manipulation, and access techniques that are
specic to particular types of data; 2) new software techniques that
can work with digital data in general (i.e. they are not specic to
particular types of data). My second scheme also divides software
techniques for working with media data into two types, but it
does this in a diferent way. What matters here are the relations
between software techniques and pre-digital media technologies.
In this taxonomy, some techniques are simulations of pre-digital
media techniques augmented with new properties and functions;
other techniques do not have any obvious equivalents in previous
physical or electronic media.
While for a media historian the second scheme is quite meaningful,
what about users who are digital natives? These software users
may never have directly used any other media besides tablets or
9781623568177_txt_print.indd 125 29/04/2013 11:50
Photoshop Toolbox from version 0.63 (1988) to 7.0 (2002).
9781623568177_txt_print.indd 126 29/04/2013 11:50
9781623568177_txt_print.indd 127 29/04/2013 11:50
laptops, or mobile media devices (mobile phones, cameras, music
players); and they are also likely to be unfamiliar with the details
of twentieth-century cel animation, lm editing equipment, or
any other pre-digital media technology. Does this mean that the
distinction between software simulations of previously existing
media tools and new born digital media techniques has no
meaning for digital natives but only matters for historians of media
such as myself?
I think that while the semantics of this distinction (i.e. the
reference to previous technologies and practices) may not be
meaningful to digital natives, the distinction itself is something
these users experience in practice. To understand why this is the
case, let us ask if all born digital media techniques available
in media authoring software applications may have something in
commonbesides the fact that they did not exist before software.
One of the key uses of digital computers from the start was
automation. As long as a process can be dened as a nite set of
simple steps (i.e. as an algorithm), a computer can be programmed
to execute these steps without human input. In the case of
application software, the execution of any command involves
low-level automation (since a computer automatically executes a
sequence of steps of the algorithm behind the command). However,
what it is important from the users point of view is the level of
automation being ofered in the commands interface.
Many software techniques that simulate physical tools share
a fundamental property with these tools: they require a user to
control them manually. The user has to micro-manage the tool,
so to speak, directing it step-by-step to produce the desired efect.
For instance, you have to explicitly move the cursor in a desired
pattern to produce a particular brushstroke using a brush tool; you
also have to explicitly type every letter on a keyboard to produce
a desired sentence. In contrast, many of the techniques that do
not simulate anything that existed previouslyat least, not in any
obvious wayofer higher-level automation of creative processes.
Rather than controlling every detail, a user species parameters
and controls and sets the tool in motion. All generative (also called
procedural) techniques available in media software fall into this
category. For example, rather than having to create a rectangular
grid made of thousands of lines by hand, a user can specify the
width and the height of the grid and the size of one cell, and the
9781623568177_txt_print.indd 128 29/04/2013 11:50
program will generate the desired result. Another example of this
higher-level automation is interpolation of key values ofered by
animation software. In a twentieth-century animation production,
a key animator drew key frames which were then forwarded to
human in-betweeners who created all the frames in between the
key frames. Animation software automates the process of creating
in-between drawings by automatically interpolating the values
between the key frames.
Thus, although users may not care that one software tool does
something that was not possible before digital computers while
another tool simulates previous physical or electronic media,
the distinction itself between the two types is something users
experience in practice. The tools that belong to the rst type
showcase the ability of computers to automate processes; the tools
that belong to the second type use invisible low-level automation
behind the scenes while requiring users to direct them manually.
Filter > stylize > wind
Having established two sets of categories for software techniques
(media-independent vs. media-specic; simulation of the old vs.
new), let us now test them against the Photoshop commands.
Before starting, however, it is important to note once again that
the two proposed schemes are intended to serve only as provi-
sional categories. They provide one possible set of directionsan
equivalent of North, South, West and East for a map where we can
locate multiple operations of media design software. Like any rst
sketch, no matter how imprecise, this map is useful because now
we have something to modify as we go forward. Our goal is not to
try to t everything we will look at into the categories of this initial
map, but rather to discover its limitations as quickly as we can, and
make modications.
We will start with Photoshop ltersi.e. the set of commands
that appear under the Filter menu. (Note that a large proportion
of Photoshop lters are not unique to this program but are also
available in other professional image editing, video editing and
animation softwaresometimes under diferent names. To avoid
any possible misunderstanding, I will be referring to the Photoshop
versions of these commands as implemented in Photoshop CS4,
9781623568177_txt_print.indd 129 29/04/2013 11:50
with their particular options and controls as dened in this
software release.
The rst thing that is easy to notice is that the names of many
Photoshop lters refer to the techniques for image manipulation
and creation and materials that were available before the devel-
opment of media application software in the 1990spainting,
drawing and sketching, photography, glass, neon, photocopying.
Each lter is given a set of explicit options that can be controlled
with interactive sliders and/or by directly entering desired numbers.
These controls not only ofer many options but also often allow
you to set lters properties numerically by choosing a precise value
from a range.
This is a good example of my earlier point that simulations
of prior physical media augment them with new properties. In
this case, the new property is the explicit lter controls. For
example, the Palette Knife lter ofers three options: Stroke Size,
Stroke Detail, and Softness. Stroke Size can take values between
1 and 50; the other two options have similarly large ranges. (At
the same time, it is important to note that expert users of many
physical tools such as a paintbrush can also achieve many efects
not possible in its software simulation. Thus, software simulations
should not be thought of simply as improvements over previous
media technologies.)
While some of these lters can be directly traced to previous
physical and mechanical media such as oil painting and photog-
raphy, others make a reference to actions or phenomena in the
physical world that at rst appear to have nothing to do with
media. For instance, the Extrude lter generates a 3D set of blocks
or pyramids and paints image parts on their faces, while the Wave
lter creates the efect of ripples on the surface of an image.
However, if we examine any of these lters in detail, we realize
that things are not so simple. Let us take the Wind lter (located
under the Stylize submenu) as an example. This is how Photoshop
CS4s built-in Help describes this lter: Places tiny horizontal lines
in the image to create a windblown efect. Methods include Wind;
Blast, for a more dramatic wind efect; and Stagger, which ofsets
the lines in the image. We are all familiar with the visual efects
For a history of Photoshop version releases, see http://en.wikipedia.org/wiki/
9781623568177_txt_print.indd 130 29/04/2013 11:50
of a strong wind on a physical environment (for instance, blowing
through a tree or a eld of grass)but before you encountered this
lter, you probably never imagined that you can wind an image.
Shall we understand the name of this lter as a metaphor? Or
perhaps, we can think of it as an example of a conceptual blend
(which is how, according to Conceptual Blending theory, many
concepts in natural languages get formed
): wind plus image
results in a new concept actualized in the operations of the Wind
The situation is further complicated by the fact that the results
of applying the Wind lter to an image look pretty diferent
compared to what the actual wind does to a tree or a eld of grass.
However, they do look rather similar to a photograph of a real
windy scene taken with a long exposure. Therefore, we can think
of the name Wind both as a metaphorto help us imagine what
a particular algorithmic transformation does to an imageand as a
simulation of a particular photographic technique (long exposure).
In short, although its name points to the physical world, its actual
operations may also refer to a pre-digital media technology.
Are there born digital lters?
Let us continue the exploration of Photoshop lters. The great
majority of the lters make references to previous physical media
See Mark Turner and Gilles Fauconnier, The Way We Think. Conceptual
Blending and the Minds Hidden Complexities (New York: Basic Books 2002).
The efects of Photoshop Wind lter (version CS5.1). Left to right:
original shape, no lter; the lter with wind option applied; the lter
with blast option applied.
9781623568177_txt_print.indd 131 29/04/2013 11:50
or our experiences in the physical worldat least in terms of how
they are named. Only a few do not. These lters include High-pass,
Median, Reduce Noise, Sharpen, and Equalize. Are these lters
born digital? In other words, did we nally get to pure examples
of new media techniques? The answer is no. As it turns out, all
these lters are also software simulations that refer to things that
already existed before digital computers.
Although they represent a small subset of Photoshops extensive
lter collection, these lters are central to all electronics, telecom-
munication, and IT technologies. Moreover, they are not unique to
processing digital images but can be used on any kind of data
sounds, television transmission, data captured by an environmental
sensor, data captured by a medical imaging devices, etc.
In their Photoshop implementation, these lters work on
continuous-tone images, but since they can be also applied to sound
and other types of signals, they actually belong to our media-
independent category of software techniques. In other words,
they are general techniques developed rst in engineering and later
also in computer science for signal and information processing.
The application of these techniques to images forms the part of
the eld of image processing dened as any form of information
processing for which the input is an image, such as photographs or
frames of video.
This conceptual relationship between infor-
mation processing and image processing exemplies one of the
key points of this bookin software culture, digital media is a
particular subset of the larger category information. (Thus, the
operations commonly used with media form a subset of a larger
category data processing.)
Like these lters, many of the new techniques for media
creation, editing, and analysis implemented in software applica-
tions were not developed specically to work with media data.
Rather, they were created for signal and information processing in
generaland then were either directly carried over to, or adapted
to work with media. (Thus, development of software brings
diferent media types closer together because the same techniques
can be used on all of them. At the same time, media now share
9781623568177_txt_print.indd 132 29/04/2013 11:50
a relationship with all other information types, be they nancial
data, patient records, results of a scientic experiments, etc.)
This is one of the most important theoretical dimensions in
the shift from physical and mechanical media technologies to
electronic media and then digital software. Previously, physical
and mechanical media tools were used to create content which
was directly accessible to human senses (with some notable excep-
tions like Morse code)and therefore the possibilities of each tool
and material were driven by what was meaningful to a particular
human sense. A paintbrush could create brushstrokes that had
color, thickness, and shapeproperties directly speaking to human
vision and touch. Similarly, the settings of photographic camera
controls afected the sharpness and contrast of captured photos
characteristics meaningful to human vision. A diferent way to
express this is to say that the message was not encoded in any
way; it was created, stored, and accessed in its native form. So if
we were to redraw the famous diagram of a communication system
by Claude Shannon (1948)
for the pre-electronics era, we would
have to delete the encoding and decoding stages.
Successive media technologies based on electronics (such as
the telegraph, telephone, radio, television), and digital computers
employ the coding of messages or content. And this, in turn,
makes possible the idea of informationa disembodied, abstract
and universal dimension of any message separate from its content.
Rather than operating on sounds, images, video, or texts directly,
electronic and digital devices operate on the continuous electronic
signals or discrete numerical data. This allows for the denition of
various operations that work on any signal or any set of numbers
regardless of what this signal or numbers may represent (images,
video, student records, nancial data, etc.). Examples of such
operations are modulation, smoothing (i.e., reducing the difer-
ences in the data), and sharpening (exaggerating the diferences).
If the data is discrete, this allows for various additional operations
such as sorting and searching.
The introduction of the coding stage allows for a new level of
efciency and speed in processing, transmitting, and interacting
with media data and communication contentand this is why rst
C. E. Shannon, A Mathematical Theory of Communication, Bell System
Technical Journal 27 (July 1948): 379423; (October 1948): pp. 62356.
9781623568177_txt_print.indd 133 29/04/2013 11:50
electronics and later digital computers gradually replaced all other
media-specic tools and machines. Operations such as those just
mentioned are now used to automatically process the signals and
data in various waysreducing the size of storage or bandwidth
needed, improving quality of a signal to get rid of noise, and
of courseperhaps most importantlyto send media data over
communication networks.
The eld of digital image processing began to develop in the
second half of the 1950s when scientists and the military realized
that digital computers could be used to automatically analyze and
improve the quality of aerial and satellite imagery collected for
military reconnaissance and space research. (Other early applica-
tions included character recognition and wire-photo standards
) As a part of its development, the eld took the basic
lters that were already commonly used in electronics and adapted
them to work with digital images. The Photoshop lters that
automatically enhance image appearance (for instance, by boosting
contrast, or by reducing noise) come directly from that period (late
1950s to early 1960s).
In summary, Photoshops seemingly born digital (or
software-native) lters have direct physical predecessors in
analog lters.

These analog lters were rst implemented by the
inventors of telephone, radio, telephone, electronic music instru-
ments, and various other electronic media technologies during the
rst half of the twentieth century. They were already widely used
in the electronics industry and studied in the eld of analog signal
processing before they were adapted for digital image processing.
Filter > distort > wave
The challenges in deciding in what category to place Photoshop
lters persist as we continue going through the Filter menu. The
two schemes for classifying software techniques for working with
Even though image processing represents an active area of computer science,
and is widely used in contemporary societies, I am not aware of any books or
even articles that trace its history. The rst book on image processing was Azriel
Rosenfeld, Picture Processing by Computer (New York: Academic Press, 1969).
9781623568177_txt_print.indd 134 29/04/2013 11:50
media I proposed turn out to be exactly what I suggested them to
beonly an initial rough map to start a discussion.
The difculties in deciding where to place this or that technique
are directly related to the history of digital computers as simulation
machines. Every element of computational media comes from
some place outside of digital computers. This is true not only for
a signicant portion of media editing techniqueslters, digital
paintbrushes and pencils, CAD tools, virtual musical instruments
and keyboards, etc.but also for the most basic computer opera-
tions such as sort and search, or basic ways to organize data such
as a le or a database. Each of these operations and structures can,
both conceptually and historically, be traced to previous physical
or mechanical operations and to strategies of data, knowledge and
memory management that were already in place before the 1940s.
For example, computer les and folders refer to their paper
predecessors already standard in every ofce. The rst commercial
digital computers from IBM were marketed as faster equivalents of
electro-mechanical calculators, tabulators, sorters and other ofce
equipment for data processing that IBM had already been selling
for decades. However, as we already discussed in detail using as
examples the work of Ivan Sutherland, and Douglas Engelbarts
and Alan Kays labs, whenever some physical operations and
structures were simulated in a computer, they were simultaneously
enhanced and augmented. This process of transferring physical
world properties into a computer while augmenting them continues
todaythink, for instance of the multi-touch interface popularized
by Apples iPhone (2007). Thus, while Alan Turing dened the
digital computer as a general-purpose simulation machine, when
we consider its subsequent development and use, it is more
appropriate to think of a computer as a simulation-augmentation
machine. The difculty of deciding how to classify diferent media
software techniques is a direct result of this paradigm that underlies
the development of what we now call software applications from
the very start (i.e. Sutherlands Sketchpad, 19623).
The shift from physical media tools and materials as algorithms
designed to simulate the efects of these tools and materials also
has another important consequence. As we saw, some Photoshop
lters explicitly refer to previous artistic media; others make
reference to diverse physical actions, efects, and objects (Twirl,
Extrude, Wind, Difuse Glow, Open Ripple, Glass, Wave, Grain,
9781623568177_txt_print.indd 135 29/04/2013 11:50
Patchwork, Pinch, and others). But in both cases, by changing
the values of the controls provided by each lter, we can vary its
visual efect signicantly on the familiar/unfamiliar dimension. We
can use the same lter to achieve a look that may indeed appear
to closely simulate the efect of the corresponding physical tool
or physical phenomenaor a look which is completely diferent
from both nature and older media and which can be only achieved
though algorithmic manipulation of the pixels. What begins as a
reference to a physical world outside of the computer if we use
default settings can turn into something totally alien with a change
in the value of a single parameter. In other words, many algorithms
only simulate the efects of physical tools and machines, materials
or physical world phenomena when used with particular parameter
settings; when these settings are changed, they no longer function
as simulations.
For an example, let us analyze the behavior of the Wave lter
(located under the Distort submenu). The lter refers to a familiar
physical phenomena, and indeed, it can produce visual efects
which we would condently call waves. This does not mean that
the efect of this lter has to closely resemble the literal meaning of
wave dened by a dictionary as a disturbance on the surface of
a liquid body, as the sea or a lake, in the form of a moving ridge
or swell.
In our everyday language, we use the word wave
metaphorically to refer to any kind of periodical movement
(waving a hand), or any static form that resembles the form of
a wave, or a disturbance in the ordinary state of afairs (making
waves.) According to an inuential theory developed by cognitive
linguist George Lakof, such metaphorical use is not an exception,
but the norm in human language and thinking. Lakof proposed
that the majority of our abstract concepts are metaphorical projec-
tions from sensorimotorial experiences with our own body and
the surrounding physical world.
Making waves and other
metaphors derived from our perceptual experience of seeing real
waves exemplify this general mechanism of language.
If we follow Lakof s theory of metaphor, some details of the
Wave lter operationalong with many other Photoshop lters
http://dictionary.reference.com/browse/wave (August 1, 2012).
George Lakof and Mark Johnson, Metaphors We Live By (Chicago: University
of Chicago Press, 1980).
9781623568177_txt_print.indd 136 29/04/2013 11:50
that refer to the physical worldcan be understood as similar
metaphorical projections. Depending on the choice of parameter
values, this lter can either produce efects that closely resemble
our perceptual experience of actual physical waves, or new efects
that are related to such waves metaphorically.
The lter generates sine wave functions (y = sin x), adds them
up and uses the result to distort an image. A user can control
the number of sine waves via a parameter called Number of
Generators. If this number is set to 1, the lter generates a single
sine wave. Applying this single function to an image distorts it
using a periodically varying pattern. In other words, the lter
indeed generates an efect that looks like a wave.
However, at some point the metaphorical connection to real
world waves breaks, and using Lakofs theory no longer works.
If we increase the number of generators (it can go up to 999), the
pattern produced by the lter no longer appears to be periodical,
and therefore it no longer can be related to real waves even
The reason for this lter behavior lies in its implementation. As
we have already explained, when the number of generators is set
to 1, the algorithm generates a single sine function. If the option
is to set to 2, the algorithm generates two functions; if it is set to
3, it generates three functions, and so on. The parameters of each
function are selected randomly within the user specied range.
If we keep the number of generators small (25), sometimes
these random values add up to a result that still resembles a wave;
in other cases they do not. But when the number of functions is
increased, the result of adding these separate functions with unique
random parameters never looks like a wave.
Photoshops Wave lter (version CS5.1). From left to right: the original
shape, no lter; the lter with 1 wave generator; the lter with 10 wave
generators; the lter with 50 wave generators.
9781623568177_txt_print.indd 137 29/04/2013 11:50
Wave lters can create a practically endless variety of abstract
patternsand most of them are not periodic in an obvious way,
i.e. they are no longer visually recognizable as waves. What they
are is the result of a computer algorithm that uses mathematical
formulas and operations (generating and adding sine functions)
to create a vast space of visual possibilities. So although the lter
is called Wave, only a tiny part of its space of possible patterns
corresponds to the wave-like visual efects in the real world.
The same considerations apply to many other Photoshop lters
that make references to physical media. Similar to the Wave lter,
the lters gathered under the Artistic and Texture submenus
produce very precise simulations of the visual efects of physical
media with a particular range of parameter settings; but when the
parameters are outside of this range, these lters generate a variety
of abstract patterns.
The operation of these Photoshop lters has important theoretical
consequences. Earlier I pointed out that the software tools that
simulate physical instrumentspaint brushes, pens, rulers, erasers,
etc.require manual control, while the tools that do not refer
to any previous media ofer higher level automation. A user sets
the parameter values and the algorithm automatically creates the
desired result.
The same high-level automation underlies generative (or
procedural) software techniques commonly used today. This work
generated by algorithms ranges from live visuals and animations by
software artist Lia
to the massive procedurally generated world in
the videogame Minecraft. Other generative projects use algorithms
to automatically create complex shapes, animations, spatial forms,
music, architectural plans, etc. (A good selection of interactive
generative works and generative animations can be found at http://
www.processing.org/exhibition/). Since most artworks created
with generative algorithms are abstract, artists and theorists like to
oppose them to software such as Photoshop and Painter that are
widely used by commercial illustrators and photographers in the
service of realism and guration. Additionally, because these appli-
cations simulate older manual models of creation, they are also
seen as less new media-specic than generative software. The
9781623568177_txt_print.indd 138 29/04/2013 11:50
end result of both of these critiques is that software that simulates
old media are thought to be conservative, while generative
algorithms and artworks are presented as progressive because these
are unique to new media. (When people claim that artworks
that involve writing computer code qualify as digital art while
artworks created using Photoshop or other media applications do
not, they rehearse a version of the same argument.)
However, as they are implemented in media applications such as
Photoshop, the software techniques that simulate previous media
and the software techniques that are explicitly procedural and use
higher-level automation are part of the same continuum. As we saw
with the Wave lter, the same algorithm can generate an abstract
image or a realistic one. Similarly, particle systems algorithms are
used by digital artists and motion graphics designers to generate
abstract animations; the same algorithms are also widely used in
lm production to generate realistic-looking explosions, reworks,
ocks of birds and other physical natural phenomena. In another
example, procedural techniques often used in architectural design
to create abstract spatial structures are also used in video games to
generate realistic 3D environments.
History and actions menus
I started this discussion of Photoshop lters to test the usefulness
of two schemes for classifying the seemingly endless variety of
techniques available in media software: 1) media-independent
vs. media-specic techniques (rst scheme); 2) the simulations
of previous tools vs. techniques which do not explicitly simulate
prior media (second scheme). The rst scheme draws our attention
to the fact that all media applications share some genes, so to
speak, while also providing some techniques that can only work
on particular data types. The second scheme is useful if we want
to understand the software techniques in terms of their genealogy
and their relation to previous physical, mechanical, and electronic
Although the previous discussion highlighted the difcult
borderline cases, in other cases the divisions are clear. For example,
the Brush Strokes lter family in Photoshop clearly takes inspi-
ration from earlier physical media tools, while Add Noise does not.
9781623568177_txt_print.indd 139 29/04/2013 11:50
The Copy and Paste commands are examples of media-independent
techniques; Auto Contrast and Replace Color commands are
examples of media-specic techniques.
However, beyond these distinctions suggested by the two schemes
I proposed, all software techniques for media creation, editing, and
interaction also share some additional common traits that we have
not discussed yet. Conceptually, these traits are diferent from
common media-independent techniques such as copy and paste.
What are they?
Regardless of whether they refer to some pre-existing instrument,
action, or phenomena in the physical world or not, media techniques
available in application software are implemented as computer
programs and functions. Consequently, they follow the principles
of modern software engineering in general. Their interfaces use
established conventions employed in all application software
regardless of whether these tools are part of spreadsheet software,
inventory management software, nancial analysis software, or
web design software. The techniques are given extensive numerical
controls; their settings can be saved and retrieved later; their use is
recorded in a History window so it can be recalled later; they can
be used automatically by recording and playing Actions; and so on.
(The terms History palette and Actions refer to Photoshop,
but the concepts behind them are found in many other software
applications.) In other words, they acquire the full function-
ality of the modern software environmentfunctionality that is
signicantly diferent from that of physical tools and machines
that existed previously. Because of these shared implementation
principles, all software applications are like species that belong to
the same evolutionary family, with media software occupying a
branch of the tree.

The pioneers of media software aimed to extend the properties of
media technologies and tools they were simulating in a computer
in each case, as formulated by Kay and Goldberg, the goal was
to create a new medium with new properties. Consequently,
software techniques that refer to previous physical, mechanical,
or electronic tools and creative processes are also new media
Contemporary biology no longer uses the idea of an evolutionary tree; the
species concept has similarly proved to be problematic. Here I am using these
terms only metaphorically.
9781623568177_txt_print.indd 140 29/04/2013 11:50
because they behave so diferently from their predecessors. We
now have an additional reason to support this conclusion. New
functionality (for instance, multiple zoom levels), the presence
of media-independent techniques (copy, paste, search, etc.) and
standard interface conventions (such as numerical controls for every
tool, preview option, or commands history) further separate even
the most realistic media simulation tool from its predecessor.
This means that to use any media authoring and editing
software is to use new media. Or, to unfold this statement:
all media techniques and tools available in software applications
are new mediaregardless of whether a particular technique
or program refers to previous media, physical phenomena, or a
common task that existed before it was turned into software. To
write using Microsoft Word is to use new media. To take pictures
with a digital camera is to use new media. To apply the Photoshop
Clouds lter (Filters > Render > Clouds) that uses a purely
automatic algorithmic process to create a cloud-like texture is to
use new media. To draw brushstrokes using the Photoshop brush
tool is to use new media.
In other words, regardless of where a particular technique would
fall in our classication schemes, all these techniques are instances
of one type of technologyinteractive application software. And,
as Kay and Goldberg explained in their 1977 article quoted earlier,
interactive software is qualitatively diferent from all previous
media. Over the next thirty years, these diferences became only
larger. Interactivity; customization; the possibility to both simulate
other media and information technologies and to dene new ones;
processing of vast amounts of information in real-time; control
and interaction with other machines such as sensors; support
of both distributed asynchronous and real-time collaboration
these and many other functionalities enabled by modern software
(of course, working together with middleware, hardware, and
networks) separate software from all previous media and infor-
mation technologies and tools invented by humans.
Layers palette
For our nal analysis, we will go outside the Filter menu and
examine one of the key features of Photoshop that originally
9781623568177_txt_print.indd 141 29/04/2013 11:50
diferentiated it from many consumer media editorsthe Layers
palette. The Layers feature was added to Photoshop 3.0, released in
To quote Photoshop Help, Layers allow you to work on
one element of an image without disturbing the others.
From the
point of view of media theory, however, the Layers feature is much
more than that. It redenes both how images are created and what
an image actually means. What used to be an indivisible whole
becomes a composite of separate parts. This is both a theoretical
point, and the reality of professional design and image editing in
our software society. Any professional design created in Photoshop
is likely to use multiple layers (in Photoshop CS4, a single image
can have thousands of layers). Since each layer can always be made
invisible, layers can also act as containers for elements that poten-
tially may go into the composition; they can also hold diferent
versions of these elements. A designer can control the transparency
of each layer, group them together, change their order, etc.
Layers change how a designer or an illustrator thinks about
images. Instead of working on a single design with each change
immediately (and in the case of physical media such as paint or ink,
irreversibly) afecting this image, s/he now works with a collection
of separate elements. S/he can play with these elements, deleting,
creating, importing and modifying them, until s/he is satised with
the nal compositionor a set of possible compositions that can
be dened using Layer Groups. And since the contents and the
settings of all layers are saved in an image le, s/he can always
come back to this image to generate new versions or to use its
elements in new compositions.
The layers can also have other functions. To again quote
Photoshop CS4s online Help, Sometimes layers dont contain
any apparent content. For example, an adjustment layer holds
color or tonal adjustments that afect the layers below it. Rather
than edit image pixels directly, you can edit an adjustment layer
and leave the underlying pixels unchanged.
In other words,
the layers may contain editing operations that can be turned on
and of, and re-arranged in any order. An image is thus redened
as a provisional composite of both content elements and various
http://help.adobe.com/en_US/Photoshop/11.0/ (October 9, 2011).
9781623568177_txt_print.indd 142 29/04/2013 11:50
modication operations that are conceptually separate from these
We can compare this fundamental change in the concept
and practice of image creation with a similar change that took
place in mappinga shift from paper maps to GIS. Just as all
media professionals use Photoshop, today the majority of profes-
sional users who deal with physical spacescity ofces, utility
companies, oil companies, marketers, hospital emergency teams,
geologists and oceanographers, military and security agencies,
police, etc.use GIS systems. Consumer mapping software such
as Google Maps, Microsoft Bing Maps and Google Earth can be
thought of as very simplied GIS systems. They do not ofer the
features that are crucial for professionals such as spatial analysis.
(An example of spatial analysis is directing software to automati-
cally determine the best positions for new supermarkets based on
existing demographic, travel, and retail data.)
GIS captures, stores, analyzes, manages, and presents data
that is linked to location.
The central concept of GIS is a stack
of data layers united by common spatial coordinates. There is an
obvious conceptual connection to the use of layers in Photoshop
and other media software applicationshowever, GIS systems
work with any data that has geospatial coordinates rather than
only images positioned on separate layers. The geospatial coordi-
nates align diferent data sets together. As used by professionals,
maps constructed with GIS software may contain hundreds or
even thousands of layers. The layers representation is also used in
consumer applications such as Google Earth. However, while in
professional applications such as ArcGIS users can create their own
layered maps from any data sources, in Google Earth users can
only add their own data to the base representation of Earth that is
provided by Google and cannot be modied.
In the GIS paradigm, space functions as a media platform which
can hold all kinds of data types togetherpoints, 2D outlines,
maps, images, video, numerical data, text, links, etc. (Other types
of such media platforms commonly used today are databases, web
pages, and spaces created via 3D compositing that I will discuss
later in the book). In Photoshop the layers are still conceptually
http://en.wikipedia.org/wiki/GIS (October 9, 2011).
9781623568177_txt_print.indd 143 29/04/2013 11:50
subordinated to the nal imagewhen you are using the appli-
cation, it continuously renders all visible layers together to show
this image. So although you can use a Photoshop image as a kind
of media databasea way to collect together diferent image
elementsthis is not the intended use (you are supposed to use
separate programs such as Adobe Bridge or Aperture to do that).
GIS takes the idea of a layered representation further. When profes-
sional users work with GIS applications, they may never output a
single map that would contain all the data. Instead, users select
the data they need to work with at that moment and then perform
various operations on this data (practically, this means selecting
a subset of all data layers available). If a traditional map ofers a
xed representation, GIS, as its name implies, is an information
system: a way to manage and work with a large sets of separate
data entities linked togetherin this case, via a shared coordinate
From programming techniques to
digital compositing
What is the conceptual origin of Layers in Photoshop? Where do
Layers belong in relation to my taxonomies of software-based
media techniques? Thinking about various possible sources of
this concept and also considering how it relates to other modern
media editing techniques takes us in a number of diferent direc-
tions. First of all, layers are not specic to raster image editors such
as Photoshop; this technique is also used in vector image editors
(Illustrator), motion graphics and compositing software (After
Efects), video editors (Final Cut), and sound editors (Pro Tools).
In programs that work with time-based datasound editors,
animation and compositing programs, and video and lm editors
layers are usually referred to as channels or tracks; these
diferent terms point to particular physical and electronic media
which a corresponding digital application has replaced (analog
video switchers, multitrack audio recorders). Despite the diference
in terms, the technique functions in the same way in all these
applications: a nal composition is a result of a adding up data
(technically, a composite) stored in diferent layers/channels/tracks.
9781623568177_txt_print.indd 144 29/04/2013 11:50
Photoshop Help explains Layers in the following way:
Photoshop layers are like sheets of stacked acetate. You can
see through transparent areas of a layer to the layers below.
(Photoshop CS4 Help, About layers.
) Although not explicitly
named by this Help article, the reference here is to the standard
technique of twentieth-century commercial. Like a lm camera
mounted above the animation stand, Photoshop software is
shooting the image created through a juxtaposition of visual
elements contained on separate layers.
It is not surprising that Photoshop Layers are closely related to
twentieth-century visual media techniques such as cel animation, as
well as to various practices of pre-digital compositing such as multiple
exposure, background projection, mattes in lmmaking, and video
However, there is also a strong conceptual link between
image Layers and twentieth-century music technology. The use of
layers in media software to separate diferent elements of a visual
and/or temporal composition strongly parallels the earlier practice
of multitrack audio recording. The inventor of multitrack recording
was the guitarist Les Paul; in 1953 he commissioned Ampex to build
the rst eight-track recorder. In the 1960s multi-track recorders were
already being used by Frank Zappa, the Beach Boys, and the Beatles;
from that point on, multitrack recording became the standard
practice for all music recording and arranging.
Originally a bulky
and very expensive machine, a multi-track recorder was eventually
simulated in software and is now available in many applications.
For instance, since 2004 Apple has included the multitrack recorder
and editor GarageBand on all its new computers. Other popular
software implementations include the free application Audacity and
the professional-level application Pro Tools.
Finally, yet another lead links Layers to a general principle of
modern computer programming. In 1984, two computer scientists
Thomas Porter and Thomas Duf working for ILM (Industrial
Light and Magic, a special efects unit of Lucaslm) formally
http://help.adobe.com/en_US/Photoshop/11.0/ (October 9, 2011).
The chapter Compositing in The Language of New Media presents an
archeology of digital compositing that discusses the links between these earlier
technologies. Lev Manovich, The Language of New Media (The MIT Press, 2001.)
See http://en.wikipedia.org/wiki/Multitrack_tape_recorder and http://en.wikipedia.
9781623568177_txt_print.indd 145 29/04/2013 11:50
dened the concept of digital compositing in a paper presented
The concept emerged from the work ILM was
doing on special efects scenes for 1982s Star Trek II: The Wrath
of Khan. The key idea was to render each separate element with a
matte channel containing transparency information. This allowed
the lmmakers to create each element separately and then later
combine them into a photorealistic 3D scene.
Porter and Dufs paper makes an analogy between creating a
nal scene by compositing 3D elements and assembling separate
code modules into a complete computer program. As Porter and
Duf explain, the experience of writing software in this way led
them to consider using the same strategy for making images and
animations. In both cases, the parts can be re-used to make new
Experience has taught us to break down large bodies of source
code into separate modules in order to save compilation time.
An error in one routine forces only the recompilation of its
module and the relatively quick reloading of the entire program.
Similarly, small errors in coloration or design in one object
should not force recompilation of the entire image.
The same idea of treating an image as a collection of elements that
can be changed independently and re-assembled into new images is
behind Photoshop Layers. Importantly, Photoshop was developed
at the same place where the principles of digital compositing were
dened earlier. Brothers Thomas and John Knoll wrote the rst
version of the program when Thomas took a six-month leave from
the PhD program at the University of Michigan in 1988 to join his
brother who was then working at ILM.
This link between a popular software technique for image
editing and a general principle of modern computer programming
is very telling. It is a perfect example of how all elements of the
modern media software ecosystemapplications, le formats,
interfaces, techniques, tools and algorithms used to create, view,
edit, and share media contenthave not just one but two parents,
Thomas Porter and Tom Duf, Compositing Digital Images, Computer
Graphics, vol. 18, no. 3 (July 1984): p. 2539.
9781623568177_txt_print.indd 146 29/04/2013 11:50
each with their own set of DNA: media and cultural practices on
the one hand, and software development on the other.
In short, through the work of many people, from Ivan Sutherland
in early 1960s, to the teams at ILM, Macromedia, Adobe, Apple
and other companies in the 1980s and 1990s, media becomes
softwarewith all the theoretical and practical consequences such
a transition entails. This section dives into Photoshops Filter and
Layers menus to discuss some of these consequencesbut more
still remain to be uncovered.
There is only software
What exactly is new media and how is it diferent from old
media? Academics, new media artists, and journalists have been
writing extensively about this question since the early 1990s. In
many of these discussions, a single term came to stand for the
whole range of new technologies, new expressive and communi-
cative possibilities, new forms of community and sociality that were
emerging around computers and Internet. The term is digital. It
received its ofcial seal of approval, so to speak, in 1996 when
the director of MIT Media Lab Nicholas Negroponte collected his
Wired columns into the book that he named Being Digital.
years later, the term digital still dominates both popular and
academic understanding of what new media is about.
When I did Google searches for digital, interactive, and
multimedia on August 28, 2009, the rst search returned 757
million results; the other two only returned 235 and 240 million
respectively. Making searches on Google Scholar produced similar
results: 10,800,000 for digital, 4,150,000 for web, 3,920,000
for software, 2,760,000 for interactive, 1,870,000 for multi-
media. Clearly, Negroponte was right: we have become digital.
I do not need to convince anybody today about the transform-
ative efects the Internet, the web, and other technological networks
have already had on human culture and society. However, what I
do want to convince you of is the crucial role of another part of
the computer revolution that has been less discussed. And yet, if we
Nicholas Negroponte, Being Digital (Vintage, 1996).
9781623568177_txt_print.indd 147 29/04/2013 11:50
really want to understand the forms of contemporary media and
also what media means today, this part is crucial. The part in
question is software.
None of the new media authoring and editing techniques we
associate with computers are simply a result of media being
digital. The new ways of media access, distribution, analysis,
generation, and manipulation all come from software. Which also
means that they are the result of the particular choices made by
individuals, companies, and consortiums who develop software
media authoring and editing applications, compression codecs,
le formats, programming and scripting languages used to create
interactive and dynamic media such as PHP and JavaScript. Some
of these choices determine conventions and protocols which dene
modern software environments: for instance, cut and paste
commands built into all software running under the Graphical User
Interface and its newer versions (such as iPhone OS), or one-way
hyperlinks as implemented in World Wide Web technology. Other
choices are specic to particular types of software (for instance,
illustration programs) or individual software packages.
If particular software techniques or interface metaphors which
appear in one applicationbe it a desktop program, web appli-
cation, or mobile appbecome popular with its users, they may
often soon appear in other apps. For example, after Flickr added
tag clouds to its interface, they soon appeared on numerous other
websites. The appearance of particular techniques in applications
can also be traced to the economics of the software industryfor
instance, when one software company buys another company, it
may merge its existing package with the software from the company
it bought. For instance, in 1995 Silicon Graphics bought two 3D
computer graphics suitesWavefront and Aliasand merged them
into a new product called Alias|Wavefront. Big companies such as
Google and Facebook are periodically buying smaller companies
and then adding the software products these companies develop to
their own oferings. Thus, one of Googles most popular applica-
tions, Google Earth is based on software originally developed by
Keyhole, Inc. and acquired by Google in 2004.
Often, techniques created for one purpose later migrate into
another area, as happened when image processing techniques made
their way into Photoshop in the late 1980s. These techniques,
developed in the second half of the 1950s for the analysis of
9781623568177_txt_print.indd 148 29/04/2013 11:50
reconnaissance photographs, are now used to creatively modify
images and to make photographs more artistic looking.
All these software mutations and new species of software
techniques are deeply socialthey do not simply come from
individual minds or from some essential properties of a digital
computer or a computer network. They come from software
developed by groups of people, marketed to large numbers of users,
and then constantly rened and expanded to stay competitive in
relation to other products in the same market category. (Google
and Facebook update their code a few times a day; GitHub, the
popular software hosting services, updates its code dozens of times
a day.)
In summary: the techniques, the tools, and the conventions of
media software applications are not the result of a technological
change from analog to digital media. The shift to digital
enables the development of media softwarebut it does not
constrain the directions in which it already evolved and continues
to evolve. They are the result of intellectual ideas conceived by
the pioneers working in larger labs, the actual products created
by software companies and open source communities, the cultural
and social processes set up when many people and companies start
using it, and software market forces and constraints.
This means that the terms digital media and new media do
not capture very well the uniqueness of the digital revolution. (I
like the term media computinghowever it is not used widely
apart from some communities in computer science primarily in
Europe). Why do they not work? Because all the new qualities of
digital media are not situated inside the media objects. Rather,
they all exist outsideas commands and techniques of media
viewers, authoring software, animation, compositing, and editing
software, game engine software, wiki software, and all other
software species. While digital representation makes it possible
for computers to work with images, text, 3D forms, sounds and
other media types in principle, it is the software that determines
what we can do with them. So while we are indeed being digital,
the actual forms of this being come from software.
Accepting the centrality of software puts in question another
fundamental concept of aesthetic and media theorythat of the
properties of a medium. What does it mean to refer to a digital
medium as having properties? For example, is it meaningful
9781623568177_txt_print.indd 149 29/04/2013 11:50
to talk about unique properties of digital photographs, electronic
texts, or websites? In their article, Kay and Goldberg do pair the
words properties and medium together: It [electronic text]
need not be treated as a simulated paper book since this is a new
medium with new properties. I have also frequently used this
combination of wordsbut it is now time to ask if it is only an
alias that can point us to a more precise concept.
Strictly speaking, while it is certainly convenient to talk about
properties of websites, digital images, 3D models, GIS representa-
tions, etc., it is not accurate. Diferent types of digital content do
not have any properties by themselves. What as users we experience
as properties of media content comes from software used to create,
edit, present, and access this content.
This includes all media authoring and viewing application
software made for both professionals and consumers, from
Photoshop to your mobile web browser. (It also includes custom
software developed for particular products such as a DVD menu
or an interactive kiosk.) So while I will continue to use the term
properties as a shortcut, you should always remember that it
stands for software techniques dened to work on particular types
of media ecologies, content and media data. (Flickrs whole system
for uploading, tagging, organizing, commenting, and sharing
images is an example of media ecology; a raster 24-bit image
stored in JPEG format is an example of a type of media data.)
It is important to make clear that I am not saying that today
all the diferences between media typescontinuous tone images,
vector images, simple text, formatted text, 3D models, anima-
tions, video, maps, music, etc.are completely determined by
application software. Obviously, these media data types have
diferent representational and expressive capabilities; they can
produce diferent emotional efects; they are processed by diferent
sensors and networks of neurons in the brain; and they are likely to
correspond to diferent types of mental processes and mental repre-
sentations. These diferences have been discussed for thousands of
yearsfrom ancient philosophy and classical aesthetic theory to
modern art and contemporary neuroscience. What I am arguing in
this book is something else. Firstly, interactive software adds a new
set of operations which can be applied to all media typeswhich
we as users experience as their new properties. (The examples
include the ability to display the same data structure in diferent
9781623568177_txt_print.indd 150 29/04/2013 11:50
ways, hyperlinking, visualization, searchability, and ndability.)
Secondly, the properties of a particular media type can vary
dramatically depending on the software application used for its
authoring and access.
Let us go though one example in detail. As an example of a
media type, we will use a photograph. In the analog era, once a
photograph was printed, all the information was xed. Looking
at this photograph at home, in an exhibition, or in a book did not
afect this information. Certainly, a photographer could produce a
diferent print with a higher or a lower contrast or use a diferent
paperbut this resulted in a physically diferent object, i.e., a
new photographic print that contained diferent information. (For
example, some details were lost if the contrast was increased.)
So what happens with a digital photograph? We can take a
photo with a dedicated digital camera or capture it with a mobile
phone, or scan it from an old book. In every case, we end up with
a le that contains an array of pixels which hold color values, and
a le header that species image dimensions, color prole, infor-
mation about the camera and shot conditions such as exposure,
and other metadata. In other words, we end up with what is
normally called digital mediaa le containing numbers which
represent the details of some scene or an object.
However, unless you are a programmer, you never directly
deal with these numbers. Instead, most of us interact with digital
media les via some application software. And depending on
which software you use, what you can do with a particular digital
media le can change dramatically. MMS (multimedia messaging
service) software on your phone may simply display a photo sent
by a friendand allow you to forward it to somebody elsebut
nothing else.
Free media viewers/players that run on desktops or on mobile
platforms typically give you more functions. For instance, a
desktop version of Googles Picasa 3.0 includes crop, auto color,
red eye reduction, variety of lters (soft focus, glow, etc.) and a
number of other functions. It can also display the same photo as
color or black and whitewithout modifying the actual digital
media le.
Finally, if I open the same photo in Photoshop, I can instruct
Photoshop to automatically replace some colors in a photo with
others, make visible its linear structure by running an edge
9781623568177_txt_print.indd 151 29/04/2013 11:50
detection lter, blur it in a dozen of diferent ways, composite with
another photo, and perform hundreds of other operations.
To summarize this discussion, let me make a bold statement.
There is no such thing as digital media. There is only software
as applied to media (or content). Or, to put this diferently: for
users who only interact with media content through application
software, the properties of digital media are dened by the
particular software as opposed to solely being contained in the
actual content (i.e., inside digital les).
Digital media is a result of the gradual development and
accumulation of a large number of software techniques, algorithms,
data structures, and interface conventions and metaphors. These
techniques exist at diferent levels of generality ranging from
a small number of very general (media-independent) ones
to thousands of very particular ones designed to do particular
tasksfor example, algorithms used to generate natural-looking
landscapes or software which can extract the camera position from
live action footage in order to correctly align a 3D model when it
is composited with this footage.
Because of the multiplicity and variety of these software
techniques, it is unwise to try to reduce digital media to a small
set of new properties. Such reduction would only be possible
if we could organize all these techniques hierarchically, seeing
them as diferent applications of a few general principles. After
thinking on and of about this for ten years (starting with my 1999
article Avant-Garde as Software where I rst tried to provide
a taxonomy of these new techniques) I eventually came to the
conclusion that such hierarchy will only mislead us. The reason is
that all these techniques equally change the identity of any media
type to which they are applied.
The fact that one technique may appear in many software
packages designed to work with diferent media types (what I
called media-independent techniques) while another technique
may be specic to a particular type of media (I called these
techniques media-specic) does not make the latter any less
theoretically important than the former. For instance, because
a zoom function is present in word processors, media viewers,
animation software, 3D modeling software, web browsers, etc.,
this does not make it more important than the algorithm designed
to do only one particular thing in relation to one media typefor
9781623568177_txt_print.indd 152 29/04/2013 11:50
instance, a spherize command which modies coordinates of all
the points in a 3D polygonal model so it appears more spherical.
I do not think that we can qualitatively measure the practical
efects on cultural production of both types of operations in this
example to conclude that one is more radical than the other. Both
operations change the media they act upon qualitatively, rather
than quantitatively. They both add new qualities (or afordances)
to media which it did have before. A Word document which can
be zoomed across multiple scales to reveal many pages at once has
a diferent media identity from one which cannot. Similarly, the
ability to precisely spherize a 3D model is a new way of working
with a spatial form that did not exist before 3D software.
In Avant-Garde as Software I grouped all new techniques of
digital media into four types based on what functions they support:
access, generation, manipulation, and analysis. But even such
simple diferentiation appears problematic to me todaypartly
because of the evolution of software since 1999, which led to
a gradual integration of these functions. For example, when a
user selects a media le on his/her laptop, tablet or a phone, the
le automatically opens in a media player/viewer program. And
today most media viewers and players (Windows Media Player,
Apples QuickTime Player, etc.) ofer some basic editing functions.
Therefore, in practical terms today you cannot simply access
media without automatically being ofered some ways to modify
it. (To be clear, I am talking here about personal computers and
mobile devices and not specialized hardware specically designed
to ofer only access, and to prevent modication of commercial
digital contentsuch as DVD players or MP3 players.)
How did we arrive at this new situation where instead of looking
at or reading content directly, most of us always experience it
through the layer of applications? The seemingly obvious answer
is the adoption of numerical code as the new universal inter-
mediary. I call it intermediary because in order to make media
accessible to our senses, it has to be analoga travelling wave
of oscillating pressure which we experience as sound, the voltage
levels applied to the pixel elements of an LCD which makes them
appear as diferent colors, diferent amounts of dyes deposited on
paper by dye-sublimation printers, and so on. Such conversions
from A to D (analog to digital) and D to A (digital to analog)
are central for digital media functioning: for example, from light
9781623568177_txt_print.indd 153 29/04/2013 11:50
waves to numbers stored in a le representing the image, and then
back to the voltage levels controlling the display. Or, in another
example, when we design an object to be printed on a 3D printer,
an analog representation on the screen is translated by a computer
into a digital le that then drives the analog signals controlling the
The two levels of encodingrst, a sampling of a continuous
analog signal to create its representation using a scale of discrete
numbers (for example, 256 levels commonly used to represent
grey tones in images), followed by a translation of this discrete
representation into a binary numerical systemmake media
incomprehensible for direct observation. The main reason for this
is not the binary code per se (invented by the Indian scholar Pingala
between the fth and second century ) since it is possible to learn
how to convert in your head a binary notation into a decimal one.
The problem is that representing even one image digitally requires
lots of numbers. For example, an image in HD resolution (1920
1080) contains 2,073,600 pixels, or 6,220,800 distinct RGB
valuesmaking it very hard to comprehend the patterns such a
set of numbers may represent if you examine them directly. (In
passing: because of these considerations, any digital image can
be understand as information visualizationrevealing patterns
contained in its numerical representation.)
Because looking at such sets of numbers with our bare eyes is
meaningless, we need to employ some technologies to translate
them into analog representations acceptable to our senses. Most
often, an image le is translated by digital hardware and software
into an image appearing on our screen. However, a digital repre-
sentation of one type of media can also be translated into another
media type that is meaningful to our senses. For example, in audio-
visual performances software often uses video to drive sound, or
reversely uses sound to generate abstract visuals. (Interestingly, the
precursor to Edisons 1877 phonographthe rst device to record
and reproduce soundwas douard-Lon Scott de Martinvilles
1857 phonautograph that transcribed sound into a visual media.
In other words, sound visualization was invented before sound
recording and reproduction.)
From the beginning, technologies that generated and transmitted
electro-magnetic analog signals (e.g, the gramophone) included at
least some controls for its modication such as changing signal
9781623568177_txt_print.indd 154 29/04/2013 11:50
amplitude. The rst well-known electronic instrument invented by
Lon Theremin in 1920 turned such controls into a new paradigm
for music performance. A performer controlled amplitude (volume)
and frequency (pitch) of a sound by moving his/her hands closer or
further away from the two antennas.
Software signicantly extends this principle by including more
controls and more ways of representing the data. For example, I
can choose to display this text I am writing now in Word as an
outline, or select a Print Layout which will show me boundaries
of pages; I can choose to see footnotes or hide them; I can ask
the application to automatically summarize the text; I can change
diferent font families and sizes, and so on. Thus, while the actual
data as it is represented and stored in a computer is no longer
directly accessible to our senses, the new model of encoding and
access has other signicant advantages since the data can be
formatted in a variety of ways. This formatting can be changed
interactively; it can be also stored with the data and recalled later.
This discussion can help us to understand the relations between
earlier electro-magnetic recording and reproduction technologies,
which were developed in the last decades of the nineteenth century,
and media software developed 100 years later. (Telephone: Bell,
1875; phonograph: Edison, 1878; television: Nipkow, 1884; radio:
Fessenden, 1900.) While previous reproduction technologies such
as woodblock printing, moveable type printing, printmaking,
lithography, and photography retained the original form of media,
the media technologies of the late nineteenth century abandoned
it in favor of an electrical signal. In other words, they introduced
coding as a way to store and transmit media. Simultaneously,
these technologies also introduced a fundamentally new layer of
mediainterface, i.e. the ways to represent (format) and control
the signal. And this in its turn changes how media functionsits
properties were no longer solely contained in the data but
were now also depend on the interface provided by technology
The shift to digital data and media software 100 years later
generalized this principle to all media. With all data types now
encoded as sets of numbers, they can only be accessed by users via
software applications which translate these numbers into sensory
representations. The consequence of this is what we have already
discussed: all properties of digital media are now dened by the
9781623568177_txt_print.indd 155 29/04/2013 11:50
particular software as opposed to solely being contained in the
actual content, i.e. digital les. So what was already true for audio
recording, radio, television, and video now also applies to text,
images, and 3D.
In short: media becomes software.
I could conclude this chapter herehowever we need to do one
more thing. It would be unfair to direct all this attention to the
term digital media without also taking up another term related
to it. Today this term is both widely used and also often put into
question. The term is new media.
Since we now understand that media today is really a set of
software techniques constantly in development, this gives a new
meaning to this troubled term. Just as there is no logical limit to the
number of algorithms which can be invented, people can always
develop new software techniques for working with media. So from
this perspective, the term new media captures well this funda-
mental logic of the computer metamedium. Software-based
media will always be new as long as new techniques continue
to be invented and added to those that already exist. And while
not every one of these techniques will change signicantly how
particular media or a combination of media will function, many
This logic of permanent extendability of media software
follows the logic of software engineering as a whole. Computer scien-
tists working in the academy and software companies constantly
develop modications of the existing algorithms, apply algorithms
from one area in another area, and develop new algorithms.
This logicwhich can be also called permanent innovationis
diferent from the logic which governed the development of media
technologies of the industrial age. Take cinema, for instance, from
1890 to 1990, (i.e. until the adoption of software tools by the lm
industry). Although the construction of lm lenses, the properties
and types of lm stock, the operations of lm camera, and other
elements of lm technology changed dramatically over the course
of the twentieth century, the basic representational capacity
of this image type remained the same. From Edison to George
Lucas, lm images continued to be produced in the same way: by
focusing light reected from the objects via lens on a at plane.
This type of capture process creates particular types of images of
9781623568177_txt_print.indd 156 29/04/2013 11:50
visible reality. Objects are rendered in one-point linear perspective.
This means that the geometric properties of any scene regardless
of its content are subjected to the same type of transformation
the same mapping that preserves some properties of the visible
(perspective distortion related to distance of objects from the
observer) as opposed to other properties (the real proportions of
object sizes, for example). Additionally once a recording was made,
the geometry of the image could not be modied later.
Once software enters the picture, these constants of the lm
medium become variables. While looking no less real, lm
image can now have multiple relationships to the world being
imaged. Digital compositing allows for the seamless insertion
of 3D computer-generated models that were not present in the
original scene. Conversely, the objects that were present can be
seamlessly removed from images. Interactive virtual panoramas
which allow the user to move around the space can be constructed
automatically. In some cases, it is even possible to re-render a lm
sequence as though it was shot from a diferent point of view. And
these are just some of the new ways in which new software changes
lm identity. (All these new possibilities, of course, also apply to
9781623568177_txt_print.indd 157 29/04/2013 11:50
9781623568177_txt_print.indd 158 29/04/2013 11:50
and evolution
9781623568177_txt_print.indd 159 29/04/2013 11:50
9781623568177_txt_print.indd 160 29/04/2013 11:50
Hybridity vs. multimedia
The rst metamedium envisioned by Kay and Goldberg in
1977 has gradually become a reality in the 1990s. Beginning
with Sketchpad and extending to the latest versions of media
software, most physical media were simulated in detail and many
new properties were added to them in the process. In parallel, a
number of brand new computational media which have no physical
precedents were also inventedfor instance, interactive navigable
3D space (Ivan Sutherland), interactive multimedia (Architecture
Machine Groups Aspen Movie Map); hypertext and hypermedia
(Ted Nelson); interactive narrative lm (Graham Weinbren);
the Internet (Licklider, Bob Taylor, Larry Roberts, and others);
the World Wide Web (Tim Berners-Lee); social media services
collaborative large-scale authoring platforms
such as Wikipedia (Jimmy Wales and Larry Sanger), and so on.
New fundamental techniques for media generation, manipu-
lation, and presentation which also had no previous physical
equivalent were also developedalgorithmic generation of line
images, 3D photorealistic rending, and the constraints originally
introduced in Sketchpad. New media-specic and general (i.e.,
D. M. Boyd and N. B. Ellison, Social network sites: Denition, history, and
scholarship, Journal of Computer-Mediated Communication, 13(1) (2007), http://
9781623568177_txt_print.indd 161 29/04/2013 11:50
media-agnostic) data management techniques were introduced.
Most importantly, by the middle of the 1990s computers became
fast enough to run all these media. In short, Kays vision of a
computer as metamediuma platform housing many existing and
new mediawas realized.
So what happens next? What is the next stage in the metamedium
development? (I am using the word stage here in a logical rather
than a historical sense). This is something that, as far as I can
see, the inventors of computational mediaSutherland, Nelson,
Engelbart, Kay and all the people who worked with themdid not
write about. However, since they set up all the conditions for this
next stage, they are indirectly responsible for it.
The discussion of the computer metamedium at the end of
Kay and Goldbergs 1977 article creates the impression that it
would develop via additions, as users built new types of media
to suit their needs using the tools provided with the personal
computer. Looking at the actual development of the computer
metamedium over the following thirty years seems to conrm
this conclusion.
For instance, if we look at the use of computer media in art
which begins in the second half of the 1950s (music composition
and algorithmic image generation), by 2003 the authoritative book
Digital Art by Christine Paul already lists dozen of diferent areas.
A Wikipedia article on Collaborative software similarly lists
about a dozen program types (and of course there are dozens or
hundreds of separate products for each type).
Another Wikipedia
article on Social Software
lists twenty major types of social
media (instant messaging, text chat, groupware, blogs, etc.)and
none of them existed in practice in the early computer days of the
To continue with these examples of such additions, a typical
visual design created today with software applications may also
appear as a simple sum of previous media: for example, a pen
drawing plus oil painting plus a photograph plus collage. Looking
at the interfaces of media editing software seems to also conrm
this impression. You see endless options for modifying a document,
appearing one after another in multiple menus. As the new options
http://en.wikipedia.org/wiki/Collaborative_software (February 4, 2012).
http://en.wikipedia.org/wiki/Social_software (February 4, 2012).
9781623568177_txt_print.indd 162 29/04/2013 11:50
become availablebecause software manufacture has released
a new version, or you have purchased some plug-ins of your
ownthey appear as additions in the same menus. And certainly,
the overall number of commands in popular media applications
gradually increases over time from release to releaseadding more
techniques for authoring, editing, remixing, and creating outputs
for diferent distribution platforms.
However, these processes of addition and accumulation are not
the only ones dening evolution of the computer metamedium.
While they are certainly at work, I think that they are not in the
center of the transformationor if you like, mutationof this
metamedium (and by extension, of all modern culture created
via software) in the three decades following Kay and Goldbergs
seminal article in 1977. (I use this yearof course, only symboli-
callyto mark the completion of the rst media invention
I believe that the new period that began in the late 1970s repre-
sents a fundamentally distinct second stage in the development of
a computer metamedium, a stage that follows the rst stage of its
invention and initial practical implementation. This new stage is
media hybridization.
Once computers became a comfortable home for a large number
of simulated and new media, it is only logical to expect that
they would start creating hybrids. And this is exactly what has
been taking place at this new stage in media evolution. Both the
simulated and new media typestext, hypertext, still photographs,
digital video, 2D animation, 3D animation, navigable 3D spaces,
maps, location information and social software toolscame to
function as building blocks for many new media combinations.
Here are some examples of computational media hybrids.
Google Earth combines aerial photography, satellite imagery, 3D
computer graphics, still photography, and other media to create
a new hybrid representation which Google engineers called a
3D interface to the planet. A motion graphics sequence may
combine content and techniques from diferent media such as live
action video, 3D computer animation, 2D animation, painting,
and drawing. (Motion graphics are animated visuals that surround
us every day; the examples are lm and television titles, TV
graphics, the graphics for mobile media content, and non-narrative
parts of commercials and music videos.) A website design may
9781623568177_txt_print.indd 163 29/04/2013 11:50
blend photos, typography, vector graphics, interactive elements,
and animation. Physical installations integrated into cultural and
commercial spacesfor example, Nobel Field at the Nobel Peace
Center in Oslo by Small Design, interactive store displays for Nokia
and Diesel by Nanika, or the interactive Memory Wall at Puerta
America hotel in Madrid by Karen Finlay and Jason Bruges, and
so oncombine animations, video, computer control, and various
interfaces from sensors to touch, to create interactive spatial media

The built-in dictionary of Microsoft Word which I am using
to write this text has a few denitions for hybrid including
the following: a plant produced from a cross between plants
with diferent genetic constituents; and an animal that results
from mating of parents from two distinct species or subspecies.
These biological meanings of hybrid capture well what has been
happening to media following its softwarization in the 1980s and
1990sthat is, the systematic translation of numerous techniques for
media creation and editing from physical, mechanical, and electronic
technologies into software tools. Translated into software, media
techniques start acting like species within a common ecologyin
this case, a shared software environment. Once released into this
environment, they start interacting, mutating, and making hybrids.
If we want to relate the beginning of the new hybridization
stage to some important projects and technologies in the history
of computational media, the famous Aspen Movie Map interactive
hypermedia system developed at MIT in 19789 would qualify as
the starting point.
A precursor to Google Street View (launched
in 2007), the system combined lm of streets in Aspen, still photo-
graphs, a navigation map that featured both aerial photography
and diagrammatic drawings, and audio. The second key event is the
release of QuickTime multimedia software by Apple on December
2, 1991 as an addition to System Software 6. As Apple explained
in QuickTime 1.0: You Oughta be in Pictures technical article
(Summer 1991): The recently introduced QuickTime 1.0 makes it
easy for you to add dynamic media like video and sound into your
http://www.davidsmall.com/articles/2006/06/01/nobel-field/; http://www.nanikawa.
com/; http://www.jasonbruges.com/projects/international-projects/memory-wall
The Interactive Movie Map: A Surrogate Travel System, video, (The Architecture
Machine, 1981), http://www.media.mit.edu/speech/videos/
9781623568177_txt_print.indd 164 29/04/2013 11:50
applications and thats just the beginning.
In the next ten years,
commercial developers, engineers, designers, and independent
media artists put lots of their creative energy into exploring the
new ability of a computer to present multiple media. Thus, I think
of the 1990s as the foundational period when many fundamental
ways of combining media within the single computer platform
were inventedfollowed by the next period of commercialization
of these inventions in the 2000s (for example, Google Earth,
introduced in 2005) and adoption to mobile platforms (iPhone,
introduced in 2007, was able to play video and songs, display
photos and hybrid maps, and send MMS).
To start the discussion of media hybridity, it is important to
make it clear that I am not simply talking about something that
already has a namecomputer multimedia, or simply multi-
media. This term became popular in the 1990s to describe
applications and electronic documents in which diferent media
types exist next to each other.
Often these media typestypically text, graphics, photographs,
video, 3D scenes, and soundare situated within what looks
visually like a two-dimensional page. Thus a typical Web page of
the 2000s is an example of multimedia; so is a typical PowerPoint
presentation. Today, at least, this is the still the most common way
of structuring multimedia documents. In fact, it is built into the
workings of most media authoring application such as presentation
software or web design software. When a user of Word, PowerPoint
or Dreamweaver creates a new document, s/he is presented with
a white page ready to display typed text; other media types have to
be inserted into this page via special commands. But interfaces
for creating multimedia do not necessarily have to follow this
convention. Email and multimedia messaging use another common
paradigm for putting elements of diferent media types together
attachments. Thus, a user of a mobile phone that supports MMS
can send text messages with attachments that can include picture,
sound, video and rich (formatted) text. Yet another paradigm
persistent in digital culturefrom Aspen Movie Map (1978) to
VRML (1994) to Second Life (2003)uses 3D space as the
Apple, QuickTime 1.0: You Oughta be in Pictures (summer 1991), http://
9781623568177_txt_print.indd 165 29/04/2013 11:50
default platform with other media such as video attached to or
directly inserted into this space.
Multimedia was an important term when interactive cultural
applications, which featured a few media types, started to appear in
large numbers in the early 1990s. The development of these applica-
tions was facilitated by the introduction of the appropriate storage
media, i.e. recordable CD-ROMs in 1991, computer architectures
and le formats designed to support multiple media le formats
(Apples QuickTime, 1991) and multimedia authoring software
(the rst version of the VideoWorks software which later was
renamed Macromedia Director was released in 1985). By the middle
of the 1990s digital art exhibitions featured a variety of multimedia
projects; digital art curricula began to ofer courses in multimedia
narrative; and art museums such as the Louvre started to publish
multimedia CD-ROMs ofering tours of their collections. In the
second part of the decade multimedia took over the Web as more
and more websites began to incorporate diferent types of media.
By the end of the decade, multimedia became the default in inter-
active computer applications. Multimedia CD-ROMs, multimedia
websites, interactive kiosks, and multimedia communication via
mobile devices became so commonplace and taken for granted that
the term lost its relevance. So while today we daily encounter and
use computer multimedia, we no longer wonder at the amazing
ability of computers and computer-enabled consumer electronics
devices to show multiple media types at once.
Seen from the point of view of media history, computer multi-
media is certainly a development of fundamental importance.
Previously multimedia documents combining multiple media
typessuch as medieval illustrated manuscripts, sacred archi-
tecture, or twentieth-century cinema and televisionwere not
interactive (in the sense of particular afordances provided by inter-
active computers, rather than other interactive technologies such
as paper books) or networked. But co-existence of multiple media
types within a single document or an application is only one of the
new developments enabled by simulation of many media types in
a computer. In putting forward the term hybrid media I want to
draw attention to another, equally fundamental development that,
in contrast to multimedia, so far has not received a formal name.
Certainly, it is possible to conceive of multimedia as a particular
case of hybrid media. However, I prefer to think of them as
9781623568177_txt_print.indd 166 29/04/2013 11:50
overlapping but ultimately two diferent developments. While some
classical multimedia applications of the 1990s would qualify as
media hybrids, most will not. Conversely, although media hybrids
often feature content in diferent media, this is only one aspect
of their make-up. So what is the diference between the two? In
multimedia documents and interactive applications, content types
in multiple media appear next to each other. In a web page, images
and video appear next to text; a blog post may similarly show text,
followed by images and more text; a 3D world may contain a at
screen object used to display video. Alternatively, each element of
a multimedia message opens in its own viewer (this was the case
for MMS implementations in the phones of the 2000s). In contrast,
in media hybrids, interfaces, techniques, and ultimately the most
fundamental assumptions of diferent media forms and traditions,
are brought together resulting in new media gestalts. That is, they
merge together to ofer a coherent new experience diferent from
experiencing all the elements separately.
Another way to underline this diference is by using the metaphor
of sexual reproduction. The result of sexual reproduction is new
individuals that combine the genetic material of their parents
rather than just mechanical assemblages of parents physical parts
(which would be analogous to multimedia). Using this metaphor,
we can say that new media ofspring similarly combine the DNA
of their media parents.
A related model that can help us to grasp some aspects of
this process is that of biological evolution. This process results
in new organisms, new species, and also new building blocks of
the organisms (molecules such as DNA and proteins.) Similarly,
sometimes new media ofspring are only slightly diferent from the
ones that already exist; at other times the combinations of software
DNA produce distinct new media species. The process of media
evolution also produces new techniques for media authoring,
editing, sharing, and collaborating, new interface conventions, and
also new algorithmsthe equivalents of the new building blocks of
biological evolution.
Yet another metaphor that can help us to understand the new
stage of the media development is remix. In the process of the
computer metamedium development, diferent media types get
remixed together, forming new combinations. Parts of these combi-
nations enter into new remixes, ad innitum.
9781623568177_txt_print.indd 167 29/04/2013 11:50
Every metaphor highlights some aspects of a phenomenon
while hiding or even distorting other aspects. The metaphors of
sexual reproduction, biological evolution, and music remix work
similarly. Each has advantages and disadvantages in explaining the
second stage of computational media development. In this chapter,
I will extensively use the concepts of hybridity and evolution to
describe the new stage in media evolution. The chapter in Part 3
will work with the remix metaphor.
Since I will be invoking each of these metaphors in diferent
parts of my narrative, you may get the impression that they are
complementary parts of the same description. But this is not the
case for hybridity and biological evolution meataphors. Besides the
everyday meaning of hybridity, it is also used in evolutionary
theory in a particular way. So if we think of hybridity in that sense,
we cannot use this concept and the biological evolution model at
the same time.
Contemporary theories of biological evolution share a basic
denition of species as a pool of organisms that can enter into
sexual relations between themselves but not with other species.
Through the processes of evolution most often predicated on
geographical separation between groups of the same species, these
groups change and eventually no longer can have reproductive
relations with one another. Such groups become new species.
In contrast, an animal hybrid is the result of interbreeding
between species. Most hybrids are produced articially, although
a few have been observed in nature.
Thus, hybrids are exceptions
in the normal evolutionary process. Therefore, when I use the term
hybrid, I am relying on a more general meaning of this word
outside of biology.
I also need to make a note about my use of the biological
evolution model. I am not suggesting that computational media
(or techno-cultural development in general) indeed evolve like
biological mechanisms, and that the mechanisms of such evolution
are the same as the mechanisms of biological evolution as formu-
lated in contemporary biology. (In his 2007 book Graphs, Maps,
Trees: Abstract Models for Literary History, Franco Moretti
provides convincing explanations of why some of the key ideas
M. L. Arnold. Natural Hybridization and Evolution (New York: Oxford
University Press, 1996).
9781623568177_txt_print.indd 168 29/04/2013 11:50
of biological evolution do not t the cultural history.
) Instead,
I want to use evolutionary theory as a rich conceptual toolbox,
which can help us to think about any kind of temporal process.
Understood in this way, evolutionary theory joins other theories of
development that aim to explain physical, social, or psychological
processeseach providing its own unique concepts which give
us additional ways to conceptualize any development. Examples
include Marxs theory of social development with its concepts of
mode of production, base, superstructure, and so on; Freuds theory
of the dream work as formulated in his 1899 Interpretation of
Dreams (concepts of condensation and displacement); complex
systems theory (concepts of emergence and self-organization); the
phase change model from thermodynamics; and many others.
Having explained how I will use the concepts of hybridity
and evolution, I will now go forward with my arguments. As I
see it, media hybridity is a more fundamental reconguration of
media universe than multimedia. In both cases we see a coming
together of multiple media types. However, multimedia does not
threaten the autonomy of diferent media. They retain their own
languages, i.e. ways of organizing media data and accessing and
modifying this data. The typical use of multiple media on the
Web or in PowerPoint presentations illustrates this well. Imagine
a typical web page from the 2000s that consists of text and video
clips embedded somewhere on the page. Both text and video
remain separate on every level. Their media languages do not spill
into each other. Each media type continues to ofer us its own
interface. With text, we can scroll up and down and zoom using
the browser controls; we can change the browsing settings so it is
displayed in a diferent font. With video, we can use its interface to
play it, pause or rewind it, loop a part, and change sound volume.
In this example, diferent media are positioned next to each other
but their interfaces and techniques do not interact. This, for me, is
a typical example of multimedia.
In contrast, in hybrid media the languages of previously distinct
media come together. They exchange properties, create new struc-
tures, and interact on the deepest levels. For instance, in motion
graphics text takes on many properties which were previously
Franco Moretti, Graphs, Maps, Trees: Abstract Models for Literary History
(London Verso, 2007).
9781623568177_txt_print.indd 169 29/04/2013 11:50
unique to cinema, animation, or graphic design. To put this difer-
ently, while retaining its old typographic dimensions of font family,
size, or line spacing, text also acquires new expressive possibilities
from cinema and computer animation. As a word moves closer to
us, it can appear out of focusas though it is a physical object
shot by a twentieth-century lm camera lens. At the same time, it
can now y in a virtual space, performing physically impossible
movesas any other 3D computer graphics object. Its proportions
change depending on what virtual lens the designer has selected.
The individual letters that make up a text string can be exploded
into many small particles; and so on. In short, in the process of
hybridization, the language of typography does not stay as it is.
Instead we end up with a new metalanguage that combines the
techniques of all previously distinct languages, including that of
Another way to distinguish between multimedia and hybrid
media is by noting whether or not the original structure of media
data is afected when diferent media types are combined. For
example, when video appears in multimedia documents such as MMS
messages, emails sent in HTML format, web pages, or PowerPoint
presentations, the structure of video data does not change in any
way. Just as with twentieth-century lm and video technology, a
digital video le is a sequence of individual frames, which have the
same size and proportions. Accordingly, the standard methods for
interacting with this data type also do not challenge our idea of
what video is. Like with VCR media players of the 19801990s,
when the user selects play, the frames quickly replace each other
producing the efect of motion. Video, in short, remains video.
This is typical of multimedia. An example of how some media
structure can be reconguredthe capacity that I take as one of the
identifying features of media hybridsis provided by The Invisible
Shape of Things Past, a well-known digital cultural heritage project
about Berlins history developed by Joachim Sauter and Dirk
Lsebrink from the media company Art+Com between 1995 and
In this project, lm clips become solid objects positioned in
a virtual 3D space. Each lm object is made from individual lm
frames situated behind one another to form a 3D stack. The angles
9781623568177_txt_print.indd 170 29/04/2013 11:50
between frames and the sizes of individual frames are determined
by the parameters of the camera that originally shot the lm. We
can interact with these new lm objects as with any other objects
in 3D space, ying around using virtual camera controls.
At the same time it is still possible to see the movie, using
the frame stack as a video player. But this operation of access has
been rethought. When a user clicks on the front frame in a stack,
the subsequent frames positioned behind one another are quickly
deleted. You simultaneously see the illusion of movement as in the
twentieth century, and the virtual 3D object shrinking at the same
In this example of media restructuring, which characterizes
media hybridity, the elements that make up the original lms data
structureindividual frameshave been placed in a new congu-
ration. The old structure has been remapped into a new structure.
This new structure retains the original data and their relationship
that is, individual lm frames are still organized into a sequence.
But it also has new dimensionssize of frames and their angles. The
new structure also enables a new type of interface for movie access,
which combines virtual space attributes and cinema attributes.
I hope that this discussion makes it clear why hybrid media
is not multimedia, and why we need this new term. The term
multimedia captured the phenomenon the content of diferent
media coming togetherbut not their languages. Similarly, we
cannot use another term that has been frequently used in discus-
sions of computational mediaconvergence. The dictionary
meanings of convergence include to reach the same point and
to become gradually less diferent and eventually the same. But
this is not what happens with media languages as they hybridize.
Instead, they acquire new propertiesbecoming richer as a result.
For instance, in motion graphics, text acquires the properties
of computer animation and cinematography. In 3D computer
graphics, rendering of 3D objects can use the techniques of
traditional painting. In virtual globes such as Google Earth and
Microsoft Virtual Earth, representational possibilities and inter-
faces for working with maps, satellite imagery, 3D buildings, and
photographs are combined to create new richer hybrid representa-
tions and new richer interfaces.
In short, softwarization of old media did not lead to their
convergence. Instead, after representational formats of older
9781623568177_txt_print.indd 171 29/04/2013 11:50
The Invisible Shape of Things Past. Joachim Sauter and Dirk Lsebrink
(Art+Com), 1995. Bottom: A lm object consisting of the frames
making up a lm clip. Top: The angles between frames correspond to the
position of the camera. Next three pages: Interaction with a lm object in
Invisible Shape application.
9781623568177_txt_print.indd 172 29/04/2013 11:50
9781623568177_txt_print.indd 173 29/04/2013 11:50
9781623568177_txt_print.indd 174 29/04/2013 11:50
9781623568177_txt_print.indd 175 29/04/2013 11:50
media types, the techniques for creating content in these media
and the interfaces for accessing them were unbundled from their
physical bases and translated into software, these elements started
interacting to produce new hybrids.
This, for me, is the essence of the new stage of computer
metamedium development. The unique properties and techniques
of diferent media have become software elements that can be
combined together in previously impossible ways.
Consequently, if in 1977 Kay and Goldberg speculated that
the new computer metamedium would contain a wide range of
already existing and not-yet-invented media, we can now describe
one of the key mechanisms responsible for the invention of these
new media. This mechanism is hybridization. The techniques and
representational formats of previous physical and electronic media
forms, and the new information manipulation techniques and
data formats unique to a computer are brought together in new
In retrospect, it is perhaps not accidental that the publication
of Kay and Goldbergs 1977 text, which, for me, summarizes the
achievements of the rst stage of computational media invention,
is directly followed by a seminal project which opens up the
next stagethat of hybridization of diferent media simulated
in software. In 19789 a group of young researchers at the
Architecture Machine Group at MIT (a pre-cursor to the MIT
MediaLab) directed by Nicholas Negroponte, constructed Aspen
Movie Mapa new type of interactive application that combined a
number of media types: video clips, maps, graphics, and diagrams.
These diferent media types were connected through a new type of
hypermedia interface. The name of the applicationAspen Movie
Mapunderscored that this application was neither a map nor
a movie but a new hybrid between the two. The project opens
the new fundamental stage in the media evolution enabled by its
softwarizationthe stage of hybridity.
The evolution of a computer metamedium
I will continue exploring the metaphor of biological evolution. As
the title of Darwins On the Origin of Species (1859) makes clear,
9781623568177_txt_print.indd 176 29/04/2013 11:50
the key goal of his evolutionary theory was to explain how diferent
species develop. Darwin proposed that the underlying mechanism
was that of natural selection. In the twentieth century, biologists
added a number of other explanations (genetic drift, mutation,
). While the mechanisms responsible for the development of
the computer metamedium are certainly diferent, we can use the
basic ideas of evolutionary theoryemergence of new species over
time, with a gradual increase in their number. But even if we only
take these basic ideas, there are important diferences between
biological and media evolution.
For example, if we compare the computer metamediums devel-
opment to a biological evolution, we can think of particularly
novel combination of media types as new species.
In biological
evolution, the emergence of new species is a very slow and
gradual process, which requires many generations.
Small genetic
changes accumulate over long periods before new species emerge.
However, new media species can emerge overnightit only
requires a novel idea and some programming. Given that today
a programmer/designer can use multiple software libraries for
media manipulation, and also specialized high-level programming
languages specically designed for rapid testing of ideas and experi-
mentation (Pure Data, Processing, etc.), a talented person can
invent new species of media in a few hours.
In evolutionary biology, species are dened as groups of
organisms. In media evolution, things work diferently. Some novel
combinations of media types may appear only once or twice. For
instance, a computer science paper may propose a new interface
design; a designer may create a unique combination for a particular
design project; a lm may combine media techniques in a novel
http://en.wikipedia.org/wiki/Evolution#Mechanisms (February 6, 2012).
I am aware that not only the details, but also even most of the fundamental
assumptions underlying evolution theory continue to be actively debated by scien-
tists. In my references to evolution, I use what I take to be a few commonly accepted
ideas from evolutionary theory. While these ideas are being periodically contested
and eventually may be disproved, at present they form part of the public common
sense: a set of widely held ideas and concepts about the world.
Natural selection is the gradual, nonrandom process by which biological
traits become either more or less common in a population as a function of difer-
ential reproduction of their bearers. It is a key mechanism of evolution. http://
en.wikipedia.org/wiki/Natural_selection (February 7, 2012).
9781623568177_txt_print.indd 177 29/04/2013 11:50
way. Imagine that in each case, a new hybrid is never replicated
again. This happens quite often.
Thus, some media combinations that emerge in the course of
media evolution will not be selected. Other combinations, on
the other hand, may survive and will successfully replicate.
(Again, remember that I am evoking the biological model only
as a metaphor, and that no claims are being made that the actual
mechanisms of media evolution are similar to the mechanisms
of biological evolution.) Eventually such successful hybrids may
become the common conventions in media design; built-in features
of media authoring applications; commonly used features in
social media sites; widely used design patterns; and so on. In
other words, they may become new basic building blocks of the
computer metamedium that can now be combined with other
An example of such a successful combination of media genes
is an image map technique for web design. This technique
emerged in the middle of the 1990s and was quickly adopted in
numerous interactive media projects, games, and websites. How
does it work? A photograph, a drawing, a color background, or
any other part of a screen is divided into a few invisible parts.
When a user clicks inside one of the parts, this activates a hyperlink
connected to this part.
As a hybrid, an image map combines the technique of hyper-
linking with all the techniques for creating and editing still images.
Previously, hyperlinks were only attached to a word or a phrase of
text and they were usually explicitly marked in some way to make
them visible (typically by underlining). When designers started
attaching hyperlinks to parts of continuous images or other parts
of a web page without explicitly showing them, a new species of
media was born.
As a new species, it denes new kinds of user behavior and it
generates a new experience of media. Rather than being immedi-
ately presented with clearly marked, ready to be acted upon
hyperlinks, a user now has to explore the screen, mousing over
and clicking until s/he comes across a hyperlinked part. Rather
than thinking of hyperlinks as discrete locations inside a dead
screen, a user comes to think of the whole screen as a live inter-
active surface. On an experiential level, rather than imagining a
hyperlink as something which is either present or absent, a user
9781623568177_txt_print.indd 178 29/04/2013 11:50
may now experience it as a continuous dimension, with some parts
of a surface being more strongly hyperlinked than others.
Another example of a successful hybrid that survived and repli-
cated in the course of recent media evolution is a virtual camera
model used in 3D computer animation. Developed in the 1980s
for creating computer animation sequences for feature lms, a
virtual camera model has gradually become one of the most widely
used elements of digital media deployed in video games, virtual
environments, program interfaces, feature lms, motion graphics,
As we will see in detail in the next part of the book, the new
language of visual design (a category which includes graphic
design, web design, interface design, motion graphics, and other
design areas) that emerged in the second part of the 1990s
ofers a particularly striking example of media hybridization.
Working with software applications, a designer can combine
any of the techniques of graphic design, typography, painting,
cinematography, animation, computer animation, vector drawing,
and 3D modeling. At the same time, s/he also can use many
algorithmic techniques for generating new images and forms (such
as particle systems or procedural modeling) and transforming them
(for instance, by using lters and other digital image processing
techniques, which do not have direct equivalents in previous
physical, mechanical or electronic media). All these techniques
are easily available within a small number of media authoring
programs (Photoshop, Illustrator, Flash, Maya, Final Cut, After
Efects, various HTML editors, etc.) and they can be easily
combined within a single design.
The result is the new design language used today in a large
number of countries around the world. The new global aesthetics
celebrates media hybridity and uses it to engineer emotional
reactions, drive narratives, and shape user experiences. The ability
to combine previously incompatible techniques of diferent media
is the single common feature of millions of designs being created
yearly by professionals and students alike, seen on the web and in
print, on big and small screens, in built environments, and all other
For the analysis of a virtual camera use, see Mike Jones, Vanishing Point: Spatial
Composition and the Virtual Camera, Animation 3, no. 2 (2007): pp. 22543.
9781623568177_txt_print.indd 179 29/04/2013 11:50
Like the post-modernism of the 1980s and the web revolution
of the 1990s, the softwarization of media (the transfer of
techniques and interfaces of all previously existing media technol-
ogies to software) has attened historyin this case, the history
of modern media. That is, while the historical origins of all the
building blocks that make up the computer metamediumor a
particular hybridmay still be important in some cases, these are
now exceptions rather than the rule. Clearly, for a media historian
such as myself, the historical origins of all techniques now available
in media authoring software do matter. They may also matter for
the people encountering a particular media designbut only if
a designer chooses to foreground this. For instance, in the logo
sequence for DC Comics (Imaginary Forces, 2005) designers used
exaggerated artifacts of print and lm to evoke particular historical
periods in the twentieth century. But when we consider the actual
process of media designthe ways in which designers work to
go from a sketch or a storyboard or an idea in their head to a
nished productthese historical origins no longer matter. When
a designer opens his/her computer and starts working, it does not
matter whether the technique was originally developed as a part
of the simulation of physical or electronic media. Thus, digital
paint brushes, lters simulating various natural textures, a camera
pan, an aerial perspective, splines, and polygonal meshes, blur and
sharpen lters, particle systems, etc.all have equal status as the
building blocks for new hybrids.
Thirty years after Kay and Goldberg predicted that the new
computer metamedium would contain a wide range of already
existing and not-yet-invented media, we can see clearly that their
prediction was correct. The computer metamedium has indeed
been systematically expanding. However, as we now can see,
this expansion should not be understood as the simple addition
of more and more new mediums. While a number of new
software mediums have been invented in the fty years following
Sketchpad, their number is probably less than one dozen. The key
process in the evolution of the computer metamedium involves
innovation on a more local levelthe previous media types
simulated in software (text, sound, drawing, etc.), the techniques
for their manipulation, and new computer-native techniques
entering in new combinations, creating a much larger number of
new species.
9781623568177_txt_print.indd 180 29/04/2013 11:50
To restate this: following the rst stage where most already
existing media were simulated in software and a number of new
computer techniques for generating and editing of media were
inventedthe stage that conceptually and practically has been
largely completed by the late 1970swe entered a new period
governed by hybridization. The already simulated mediums started
exchanging properties and techniques. As a result, the computer
metamedium came to contain endless new species. In parallel,
we do indeed see a continuous process of the invention of the
newbut what are being invented are not whole new media types
but rather new elements (new techniques for creating, modifying
and sharing media data). As soon as they are invented, these new
elements start to interact with other, already existing elements.
Thus, the processes of invention and hybridization are closely
linked and work together.
This, in my view, is the key mechanism responsible for the
evolution and expansion of the computer metamedium from
the late 1980s until nowand I do not see any reason why this
mechanism would become less important in the future. And while
at the time when Kay and Goldberg were writing their article the
process of hybridization was just barely startingthe rst truly
signicant media hybrid was Aspen Movie Map project created
at MITs Architecture Machine Group in 19789today it is
what media design is all about. Thus, from the point of view of
today, the computer metamedium is indeed an umbrella for many
thingsbut rather than only containing a set of separate mediums,
it also contains a larger set of smaller building blocks that unite to
create hybrids. These building blocks include algorithms for media
creation and editing, data formats, interface metaphors, navigation
techniques, physical interaction techniques, web technologies, and
other element types. Over time, new elements are being invented and
they also become parts of the computer metamedium. Periodically
people gure out new ways in which some of the elements available
can work together, producing new species. Some of these species
may survive. Some may become new conventions, so omnipresent
that they are not perceived anymore as combinations of elements
which can be taken apart. Still others are forgottenonly to be
sometimes reinvented later.
Clearly, all the building blocks that together form the computer
metamedium do not have equal importance and equal linking
9781623568177_txt_print.indd 181 29/04/2013 11:50
possibilities. Some are used more frequently than others, entering
in many more combinations. The virtual 3D camera model is
currently more widespread than, for example, techniques for
rendering realistic looking hair or fur. The 3D camera model is
built into every 3D animation application and consequently is
used in numerous 3D animations and visual efects sequences; it
appears in TV commercials, motion graphics, instructional video,
and feature lms; it is a part of the user interface in all 3D video
games; and is also the interface to popular 3D virtual globes such
as Google Earth.
In contrast, hair and fur algorithms may not
be available in every animation package and their applications are
also more limited, since only human and animal characters can
have hair or fur.
Some of the new inventions may become so important and
inuential that it seems no longer appropriate to think of them as
just elements. Instead, they may be more appropriately called new
media platformsor simply new mediums.
Mobile media platforms which emerged in the late 2000s
iOS and Android powering both tablets and mobile phonesare
perfect examples here. 3D virtual space, the World Wide Web,
and geo media (media which includes GPS coordinates) are
other examples of such new media platforms popularized in the
1980s, 1990s, and 2000s, respectively. These media platforms
fundamentally recongure how all other media are understood
and how they can be used. Thus, when we add spatial coordi-
nates to media objects (geo media), place these objects within a
single global networked hypertext (the web), or when we start
using 3D virtual space as a new platform to design not only
buildings and industrial objects but also movies and cartoons,
the identity of what we think of as media changes in funda-
mental ways. In fact, we can even say that these changes have
been as fundamental as the efects of media softwarization in
the rst place.
But is it true? There is no easy way to resolve this question.
Ultimately, it is a matter of perspective. For instance, the simulation
of existing media in software and the subsequent period of
media hybridization has had a much more substantial impact on
For a list of other virtual globe applications and software toolkits, see http://
en.wikipedia.org/wiki/Virtual_globe (February 7, 2012).
9781623568177_txt_print.indd 182 29/04/2013 11:50
contemporary visual and spatial aesthetics across all design elds
(at least so far) than did the invention of the web and graphical web
If we are interested in the histories of visual communication,
techniques of representation, and cultural memory, I do think
that the universal adoption of software throughout global culture
industries is at least as importance as the invention of print,
photography or cinema. But if we are to focus on the social and
political aspects of contemporary media culture and ignore the
questions of how media looks and what it can representasking
instead about who gets to create and distribute media, how people
understand themselves and the world through media, and how
they create and maintain social relationswe may want to put
computer networks (be the Web of the 1990s, social media of the
2000s, and whatever new yet-to-be-invented forms will come in
the future) in the center.
And yet, it is important to remember that without software,
contemporary networks would not exist. Logically and practically,
software lies underneath everything that comes later.
For example, if I disconnect my laptop from my wireless
network right now, I can still continue using most of the applica-
tionsincluding Word to write this sentence. I can also edit images
and video, create a computer animation, design a fully functional
website, and compose blog posts. (Of course by the time you are
reading this, Microsoft may be ofering Word only as an online
service, but some other word processors which run locally should
be still available)
But if somebody disables the software running the network, it
will go dead.
In other words, without the underlying software
layers The Internet Galaxy (to quote the title of the 2001 book
by Manuel Castells
) would not be possible. And if software
was already responsible for the very rst ARPANET (Advanced
Since the late 2000s, there has been a gradual movement towards ofering more
and more functionality in web applications. However, at least today (2013), unless
I am in Singapore or Tallinn which are completely covered with free Wi-Fi courtesy
of their governments, I never know if I will nd a network connection or not, so I
would not want to completely rely on the webware.
Manuel Castells, The Internet Galaxy: Reections on the Internet, Business, and
Society (Oxford: Oxford University Press, 2001).
9781623568177_txt_print.indd 183 29/04/2013 11:50
Research Projects Agency Network) computer network which
linked two remote machines on October 29, 1969 at the Network
Measurement Center at UCLAs School of Engineering and
Applied Science and Douglas Engelbarts NLS system at SRI
International in Menlo Park, California, its importance and
variety only increased as networks develop. Thus, a myriad of
software technologies are what allows for media to exist on the
web in the rst place: images and video in web pages, blogs,
Facebook and Twitter, media sharing services such as YouTube
and Instagram, aerial photography and 3D buildings in virtual
globes, etc. Similarly, the use of 3D virtual space as a platform for
media design (which will be discussed in detail in the next part)
really means using a number of software algorithms that control
the virtual camera, position the objects in space, calculate how
they look in perspective, simulate the spatial difusion of light on
the surfaces, and so on.
Hybridity: examples
The examples of media hybrids are all around us: they can be
found in user interfaces, web applications, mobile apps, visual
design, interactive design, visual efects, locative media, interactive
environments, digital art, and other areas of digital culture. Here
are a few examples that I have deliberately drawn from diferent
areas. Created in 2004 by Stamen Design (San Francisco), Mappr
was one the rst popular web mashups.
It combined a geographic
map and photos from Flickr.
Using information entered by Flickr
users, the application guessed geographical locations where photos
were taken and displayed them on the map. (Today similar map
interfaces to photo collections that use GPS data captured with
photos are available for iPhoto, Instagram, and other photo
services and apps.) Since May 2007, Google Maps has ofered
Street Views that add panoramic photo-based views of city
streets to other media types already used in Google Maps.
http://en.wikipedia.org/wiki/Web_mashup (February 7, 2012).
http://stamen.com/projects/mappr (November 3, 2012).
http://maps.a9.com (January 27, 2006).
9781623568177_txt_print.indd 184 29/04/2013 11:50
hybrid between photography and interfaces for space navigation,
Street Views allows users to navigate through a space on a street
level by clicking on the arrows superimposed on the panoramic

Starting in 1991, Japanese media artist Masaki Fujihata created
a series of projects called Field Studies.
These projects place video
recordings made in particular places within highly abstracted
3D virtual spaces representing these places. Fujihata started to
work on Field Studies a decade before the term locative media
made its appearance. As cameras with built-in GPS did not yet
commercially exist at that time, the artist created a special video
camera which captured geographical coordinates of each interview
locationalong with the cameras direction and angle while he
was video-taping the interview, as well as his movement though
space. The artist used this captured information to create a media
interface which combined 3D navigable space and video in a
unique way.
For instance, to create the interactive installation Field-Work@
Alsace (2002),
Fujihata recorded a number of video interviews
with the people living in and passing through the area around
the border between France and Germany. The project confronts
us with a black screen, with a number of three-dimensional
white lines showing the artists movement through the area as he
was capturing the interviews. As we navigate around the space,
the changing perspective views of these lines suggest the shapes
of the Alsace terrain. We also see a number of at rectangles
that are positioned at points where each interview was recorded.
Each rectangle is situated at a unique angle that corresponded
to the angle of the hand-held video camera during the interview.
When you click on a rectangle, the corresponding video plays
In my view, Alsace represents a particularly interesting media
hybrid. It fuses photography (still images which appear inside
rectangles), video documentary (video playing once a user clicks
inside a rectangle), the locative media (the movement trajectories
recorded by GPS) and 3D virtual space. In addition, Alsace uses a
http://en.wikipedia.org/wiki/Google_Street_View (July 17, 2008).
www.field-works.net/ (January 27, 2006).
http://www.medienkunstnetz.de/works/field-work/ (February 11, 2012).
9781623568177_txt_print.indd 185 29/04/2013 11:50
Field-Work@Alsace by Masaki Fujihata, 2002.
9781623568177_txt_print.indd 186 29/04/2013 11:50
new media technique developed by Fujihata: the recording not just
of the 2D location but also of the 3D orientation of the camera.
The result is a new way to represent collective experiences
using 3D space as an overall coordinate systemrather than, for
instance, a narrative or a database. At the same time, Fujihata
found a simple and elegant way to render the subjective and
unique nature of each video interviewsituating each rectangle
at a particular angle that shows where the camera was during the
interview. Additionally, by dening 3D space as an empty void
containing only trajectories of Fujihatas movement through the
region, the artist introduced the additional dimension of subjec-
tivity. Even today after Google Earth has made 3D navigation of
space containing photos and video a common experience, Alsace
and other projects by Fujihata continue to stand out. They show
that to create a new kind of representation it is not enough to
simply add diferent media formats and techniques together.
Rather, it may be necessary to systematically question the conven-
tions of diferent media types to make up a hybrid, changing their
structure in the process.
A well-known media art project I have already evokedThe
Invisible Shape of Things Pastalso uses 3D space as an umbrella
that contains other media types. As I have already discussed, the
project maps historical lm clips of Berlin recorded throughout
the twentieth century into new spatial forms that are integrated
into a 3D navigable reconstruction of the city.
The forms are
constructed by placing subsequent lm frames behind each other.
In addition to being able to move around the space and play the
lms, the user can mix and match parts of Berlin by choosing from
a number of maps which represent city development in diferent
periods of the twentieth century. Like Alsace, Invisible Shape
recombines a number of common media types while changing their
structure. A video clip becomes a 3D object with a unique shape.
Rather than representing a territory as it existed in a particular
time, a map can mix parts of the city as they existed at diferent
Another pioneering media hybrid created by Sauter and his
company Art+Com is Interactive Generative Stage (2002)a
See www.artcom.de
9781623568177_txt_print.indd 187 29/04/2013 11:50
virtual set whose parameters are interactively controlled by actors
during the opera.
During the opera performance, the computer
reads the body movements and gestures of the actors and uses this
information to control the generation of a virtual set projected
on a screen behind the stage. The positions of a human body are
mapped into various parameters of a virtual architecture such as
the layout, texture, color, and light.
Sauter felt that it was important to preserve the constraints of
the traditional opera formatactors foregrounded by lighting
with the set behind themwhile carefully adding new dimen-
sions to it.
Therefore, following the conventions of traditional
opera the virtual set appears as a backdrop behind the actors
except now it not a static picture but a dynamic architectural
construction that changes throughout the opera. As a result, the
identity of a theatrical space changes from that of a backdrop to a
main actorand a very versatile actor at thatsince throughout
the opera it adopts diferent personalities and continues to
surprise the audience with new behaviors. This kind of funda-
mental redenition of an element making a new hybrid is rare,
but when a designer is able to achieve this, the result is very
Not every hybrid is necessarily elegant, convincing, or forward-
looking. Some of the interfaces of popular media creation and access
applications look like the work of an aspiring DJ, mixing opera-
tions from old interfaces of various media with new GUI principles
in sometimes erratic and unpredictable ways. In my view, a striking
example of such a problematic hybrid is the interface of Adobe
Acrobat version 8.0, released in November 2006.
(Note that since
the interfaces of all commercial software applications typically
change from version to version, just as elsewhere in the book, this
example refers to this particular version of Adobe Acrobat.) This
version of Acrobats User Interface combines interface metaphors
from a variety of media traditions and technologies in a way that,
at least to me, does not always seem to be logical. Within a single
interface, we get 1) the interface elements from analog media
The full name of the project is Interactive generative stage and dynamic costume
for Andr Werners opera, Marlowe, the Jew of Malta.
Joachim Sauter, personal communication, Berlin, July 2002.
http://en.wikipedia.org/wiki/Adobe_Acrobat#Version_8.0 (February 8, 2012).
9781623568177_txt_print.indd 188 29/04/2013 11:50
recorders/players of the twentieth century, e.g., VCR-style arrow
buttons; 2) interface elements from image editing software, e.g.,
a zoom tool; 3) interface elements which have strong association
with the print traditionalthough they never existed in print
(page icons also controlling the zoom factor); (4) elements which
have existed in books (the bookmarks window); (5) the standard
elements of a GUI such as search, lter, and multiple windows.
It seems that Acrobat designers wanted to give users a variety of
ways to navigate through documents. However, I nd the use of
so many navigation metaphors confusing. For instance, given that
Acrobat was designed to closely simulate the experience of print
documents, it is not clear to me why I am asked to move through
the pages by clicking on the forward and backward arrowsan
interface convention which is normally used for moving image
The hybrids do not necessarily have to involve a deep
reconguration of previously separate media languages and/or the
common structures of media objectsthe way, for example, The
Invisible Shape recongures the structure of a lm object. Consider
web mashups which combine data elements from multiple sources,
hiding this behind a simple unied graphical interface.
I have
already used one example of a Web mashupMappr. Here are
some other successful early examples. Flickrvision 3D (David Troy,
2007) used data provided by Flickr and the virtual globe from
Poly 9 FreeEarth to create a mashup which continually showed
the new photos uploaded to Flickr attached to the virtual globe
at those locations where the photos were taken. Another mashup
called Liveplasma (2005) used Amazon services and data to ofer a
music and discovery engine. When a user selected an actor, a movie
director, a movie title, or a band name, Liveplasma generated an
interactive map that showed related actors, movie directors, etc.
using various dimensions such as style, inuences, popularity, and
Although Liveplasma suggests that the purpose of these
maps is to lead you to discover the items that you are also likely
to like (so you purchase them on amazon.com), these maps are
valuable in themselves. They employ newly available rich data about
peoples cultural preferences and behavior collected by Web 2.0 sites
http://en.wikipedia.org/wiki/Mashup_(web_application_hybrid) (July 19, 2008).
http://www.liveplasma.com/ (August 16, 2008).
9781623568177_txt_print.indd 189 29/04/2013 11:50
such as Amazon to do something that was not possible until the
2000s. That is, rather than mapping cultural relationships based on
the ideas of a single person or a group of experts, they reveal how
these relationships are understood by actual cultural consumers.
The development of mashups is supported by the gradually growing
number of web APIs ofered by a variety of companies. An API provides
an easy way for a programmer to create new programs, which use
services or data provided by web companies. For example, you can use
the Google Maps API to generate interactive Google maps inside your
website. When a user comes to the site and enters an address into the
map interface, Google servers get the request and send back the new
map. When I checked the mashup tracker programmableweb.com on
February 8, 2012, it listed 2,337 mashups that use the Google Maps
Many mashups combine between half a dozen and a dozen APIs
from diferent services. By September 2012, the site tracked over 7,000
diferent APIs.
Sorted by the number of mashups using these APIs,
the top entries were the Google Maps API used in 2,416 mashups,
the Twitter API used in 717 mashups, and YouTube, used in 650
These numbers may only represent a small percentage of
all mashups out therejust think about all the times you encounter a
Google map used as part of some website or service. According to one
2012 estimate, the Google Maps API was used in 350,000 web sites.
Thus, the numbers reported by programmableweb.com perhaps only
indicate the relative proportions in APIs use in mashups.
Visually, many mashups may appear as typical multimedia web
pagesbut they are more than that. As the Wikipedia article on
mashup (web application hybrid) explains, A site that allows a
user to embed a YouTube video for instance, is not a mashup site
the site should itself access 3rd party data using an API, and process
that data in some way to increase its value to the sites users. (My
emphasis). Although the terms used by the authorsprocessing data
to increase its valuemay appear to be strictly business-like, they
also capture the diference between multimedia and hybrid media
http://www.programmableweb.com/apis/directory/1?sort=mashups (February 8,
time-last-year/ (August 23, 2012).
http://www.programmableweb.com/apis/directory/1?sort=mashups (November 4,
9781623568177_txt_print.indd 190 29/04/2013 11:50
in a theoretically accurate way. Paraphrasing the articles authors,
we can say that in the case of successful artistic hybrids such as The
Invisible Shape or Alsace, separate representational formats (video,
photography, 2D map, 3D virtual globe) and media navigation
techniques (playing a video, zooming into a 2D document, moving
around a space using a virtual camera) are brought together in ways
which increase the representational and expressive value ofered by
each media type used. However, in contrast to the web mashups
that started to appear en masse in 2006 when Amazon, Flickr,
Google and other major web companies ofered public API (i.e.,
they made it possible for others to use their services and some of
their datafor instance, using Flickr images as a part of a mashup),
these projects also used their own data, which the artists carefully
selected or created themselves. As a result, the artists have much
more control over the aesthetic experience and the personality
projected by their works than an author of a mashup, which relies
on both data and the interfaces provided by other companies and
non-prot organizations such as OpenLayers.
I am not trying to criticize the web mashup technologyI only
want to suggest that if the projects goal is to put forward a diferent
representational model and a unique aesthetic experience, choosing
from the same set of web sources and data sources available to
everybody else may be not the right solution. And the argument
that web mashup author acts as a DJ who creates by mixing what
already exists also does not work heresince a DJ has both more
control over the parameters of the mix, and more recordings to
choose from.
In discussing examples of hybrids so far I have implicitly
presented them as combinations and recongurations of previously
existing media, which include both the simulations of physical
media such as print and new media such as computer animation. In
other words, I have relied on the notion consistent with Kays own
formulation that the computer mediamedium can be understood as
a collection of diferent mediums. For example, I talked about how
Alsace combined photography and video documentary (pre-digital
media simulated in a computer) with locative data and 3D virtual
space (new computer media).
However, we can also think of media hybridity using a diferent
conceptualization of the metamedium. That is, rather than empha-
sizing whole media we can focus on their building blocksi.e.,
9781623568177_txt_print.indd 191 29/04/2013 11:50
diferent types of media data (or media content) and two types
of techniques which can operate on this data. (I called these
techniques media-specic, if they can only work on specic
media types; I called media-independent those techniques that
are implemented to work with many types).
From this perspective, the new hybrid media speciesa single
project, web service, or a software programrepresents the
meeting of various techniques that previously belonged to diferent
mediums. Chapter 5 will develop this concept in detail using
examples from motion graphics and design. We will also see how
hybridization as enabled by software became one of the dominant
aesthetics of contemporary media. But as a way of getting started,
let us take the example of hybridization and discuss it in terms of
data types and data manipulation techniques. For this example, I
will use an application which should be familiar to everybody and
which I have already briey evoked a number of timesGoogle
Google Earth is based on an earlier application called Earth
Viewer developed by Keyhole, Inc. This company was acquired
by Google in 2004. In its turn, Earth Viewer took the idea of
seamless interactive navigation cinema-style around the detailed
and hybrid spatial representation from 1996 project Terravision.
The project was created by the same innovative design team which
is also responsible for some of the other outstanding media hybrids
we already encounteredJoachim Sauter and Art+Com.
following analysis applies the features and interfaces in Google
Earth 5, released in May 2009.) Using this application, you can
navigate around the Earths surface, zooming in and out; turn on
and of a variety of data overlays; search for places and directions;
mark places on the map and share these additions with all other
users of Google Earth; import your own information including
images and GPS data; create movies of touring around; and more.
When Google Earth was rst released in June 2005, Google
called it a 3D interface to the planet.
This description itself tells
us right away that we are not dealing with a twentieth-century map
or any other representation already familiar to users. So what are
http://windowssecrets.com/langalist-plus/a-3d-interface-to-the-planet/ (February
10, 2012).
9781623568177_txt_print.indd 192 29/04/2013 11:50
the key elements of the experience ofered by this 3D interface to
the planet that make it stand out from the variety of other cultural
applications that allow a user to navigate around and perform
actions on some data2D Google Maps, web browsers, iTunes,
educational multimedia applications, and so on? These elements
are both its hybrid terrain and the corresponding hybrid navigation
The representation of Earths surface that appears in the main
Google Map window, called 3D Viewer, combines satellite
photography, 3D elevation data, 3D models of buildings, and the
graphics elements familiar to us from modern paper maps (vector
graphics and text labels identifying roads, country boundaries,
etc.). Importantly, the four types of data are glued together, (i.e.,
rendered directly on top of each), thus appearing as a single visual
source. This is a perfect example of a hybrid. The diferent media
types are brought together to create a new representation.
The 3D interface ofered by Google Earth is also a hybrid. It
draws on the new type of computational media that has been
evolving since the late 1960s: 3D interactive navigable space. It
also uses the computer simulation of Hollywood cinematography
techniques developed in the 3D computer animation eld since the
1970s. The user navigates around the hybrid terrain using a set of
dened camera controls that extend the language of zoom, tilt, and
pan, developed in lm cinematography. (Google Earth 6 denes
the following 3D navigation techniques: move left, right, up and
down, rotate clockwise and counter wise, tilt up and down, zoom
in and out, zoom + automatic tilt, look and reset.
In addition to this basic navigation system, the application also
ofers a more explicitly cinematic and more automatic method
called Touring, where the camera ies between the points in a
seamless trajectory.
(I am distinguishing between 3D interactive space and
simulated camera for the following reason. While software
used by computer animators, game designers, and media designers
provides a virtual camera interface to 3D space with all traditional
cinematographic controls, other applications which also use virtual
spaces such as VR and computer games do not. Instead they use
376154&ctx=topic (February 10, 2012).
9781623568177_txt_print.indd 193 29/04/2013 11:50
The Layers window in Google Earth 7.0 (2012). The screenshot shows
only some of the hundreds of layers available.
9781623568177_txt_print.indd 194 29/04/2013 11:50
movements of the body, head, or ngers of the human user to guide
a camera. Thus, a 3D space representation and a 3D camera model
do not have to go together.)
While Google Earths core data model (satellite imagery +
elevation data + map symbols) remains unchanged over time and
a number of software releases, the continuing addition of new
data sources and data types makes the representation increasingly
richand simultaneously increases its hybridity. These additional
types of data include links to web content, Street View (launched
on May 25, 2007), normal and panoramic hi-res photos, historical
imagery (in Google Earth 5.0), underwater terrain, the Moon,
Mars, real-time trafc, etc. The addition of some of these new data
types requires parallel addition of new navigation mechanisms.
Thus, side-by-side with the original core 3D cinema-like interface
we now nd other interfaces.
In this way, the techniques for working with data provided by
Google Earth also become more hybrid. When interacting with a
3D building, a user can swoop to the top or side of the building.
In the case of high-res gigapixel photos, Google Earth ofers a
special way of ying into a photo which then can be panned
and zoomed. And with Street View, yet another set of navigation
techniques is provided.
Strategies of hybridization
As we see, media hybrids can be structured in diferent ways. In
user interfaces such as the interface of Acrobat Reader, the opera-
tions which previously belonged to specic physical, mechanical,
and electronic media are combined to ofer the user more ways to
navigate and work with the computer documents (combination of
diferent interface techniques). Google Earth combines diferent
types of media to provide more comprehensive information about
places when either media can do by itself (a combination of media
Using the keyboard or mouse the horizontal and vertical viewing direction
and the zoom level can be selected. A solid or broken line in the photo shows the
approximate path followed by the camera car, and arrows link to the next photo
in each direction. At junctions and crossings of camera car routes, more arrows are
shown. http://en.wikipedia.org/wiki/Google_Street_View (February 11, 2012).
9781623568177_txt_print.indd 195 29/04/2013 11:50
types). Mappr exemplies another strategy: using a 2D geo map as
an interface to a media collectionin this case, photos uploaded
on Flickr (using one media type as an interface to another media
type.) Alsace and Invisible Shape exemplify yet another type of
media combination: using one media type as an enclosure for
another media type (3D virtual space containing lm and video
and clips).
A complementary way of categorizing media hybrids is by asking
if a particular hybrid ofers new ways of representing the world,
and/or new ways of navigating these representations. Hybrids may
combine and/or recongure familiar media formats and media
interfaces to ofer new types of hybrid representations. For instance,
Google Earth and Microsoft Bing Maps combine diferent media
types and interface techniques to provide more comprehensive
information about places than either media can do by itself. The
ambitions behind Alsace and Invisible Shape are diferentnot to
provide more information by combining existing media formats
but rather to recongure these formats in order to create new
representations of human collective and individual experiences that
fuse objective and subjective dimensions. But in both cases, we can
say that the overall goal is to represent the world or our experience
in a new way by combining and possibly reconguring already
familiar media representations (photos, video, maps, 3D objects,
web pages, panoramic photos, etc.) Another good example of such
hybrids is Microsoft Photosynth which ofers new types of 3D
representations (synths) made by matching many photographs
of the same scenessuch as a detailed model of a 3D Notre Dame
Cathedral created entirely from its photos on Flickr.

Secondly, the hybrids may focus on new ways of navigation and
interaction with already existing media formats. Here the media
type itself is neither modied nor combined with other media
instead, hybridization happens in the UI and the tools provided by
the project, service or the application for working with this media
type. For example, in the case of Mappr, both 2D map and photo
formats already existed separately. The mashup links them together,
turning the map into an interface to the photos available on Flickr.

(February 19, 2012).
This mashup also exemplies an important development within metamedia
9781623568177_txt_print.indd 196 29/04/2013 11:50
(Flickr itself later ofered the similar map interface
as well as an
application to allow users to place their photos on a world map.

On February 19, 2012, Flickrs map interface contained over 175
million geo-tagged photos.)
In summary, a hybrid may dene new navigation and inter-
action techniques that operate over non-modied media formats.
Alternatively, a hybrid may dene new media formats but use
already existing interaction/interface techniques. A hybrid may also
combine both strategies, i.e. it can dene both new interfaces/tools
and new media formats at the same time. This, however, requires
a real creativity and deep understanding of both media computing
and media aesthetics, so such hybrids do not appear very often.
(Alsace, Invisible Shape, and Photosynth are able to combine both
strategies, and this is why for me they stand out from the multitude
of new media projects and applications created in the last two
You may notice that the distinction between a representation
(or a media format) and an interface/tool corresponds to the
two fundamental components of all modern software: data struc-
tures and algorithms. This is not accidental. Each tool ofered by a
media authoring, editing or viewing application corresponds to an
algorithm that either processes the data in a particular format in
some way, or generates new data in this format. For example, let
us assume that our media format is a photo (or, more generally, a
raster image). To generate a gallery view of the photos an algorithm
has to process each photo to t it into a specied size (this is done
by calculating averages of groups of pixels and using a new smaller
set of these average values). To draw a line over a photo requires
calling another algorithm that calculates new colors for the pixels
beneath the line. Thus, working with media using application
software essentially means running diferent algorithms over the
evolution: a convergence between media and spatial data. The three main forms of
this convergence are: 1) a 2D map used as an interface to other media types (as in
Mappr); 2) a 3D virtual map used as an interface to other media types (as in Alsace,
Invisible Shapes or Google Earth); 3) location information automatically added by
a capture device to images and video recordings.
http://en.wikipedia.org/wiki/Flickr#Organizr (March 2, 2012).
9781623568177_txt_print.indd 197 29/04/2013 11:50
While this logical diferentiation is clear and useful for the
person who understands programming, when we consider the users
experience of media authoring/viewing/cataloging/sharing applica-
tions, web services, and interactive media projects, it is harder to
maintain. In the world of application software, media data and
interfaces/tools never exist in isolation from each other. Unless
you know how to program, you never encounter media content
typesdigital photos, digital videos, maps, etc.by themselves.
Instead, you encounter media content through particular software
applications, or the custom interfaces dened by the designers of
a particular project. In other words, you always work with data
in a context of some application, one that comes with its own
interface and tools. This means that as experienced by a user of
application software, representation consists of two interlinked
parts: data structured in particular ways and the interfaces/tools
provided to navigate and work with this data. (The same applies
for the concept of information.) For example, a 3D virtual
space as it is dened in 3D computer animation and CAD appli-
cations, computer games, virtual globes, and other applications
is not only a set of coordinates that make up 3D objects and a
perspective transformation but also a set of navigation methods
i.e., a virtual camera model. A photograph as dened by media
editing applications includes various editing operations that can
be performed on it such as scale, cut and paste, make mask, add
layers, etc. Liveplasmas interactive culture maps are not only
relationships between the items on the map that we can see but
also the tools provided to construct and navigate these maps. And
the unique Earth in Google Earth is made up not only from its
hybrid data model (satellite photography, elevation, 2D map, 3D
buildings, panoramas) but also the rich techniques for navigating
and exploring this data.
9781623568177_txt_print.indd 198 29/04/2013 11:50
Soft evolution
Algorithms and data structures
What makes possible the hybridization of media creation, editing,
and navigation techniques? To start answering this question we
need to ask once again what it means to simulate physical media in
software. For example, what does it mean to simulate photography
or print media?
A nave answer is that computers simulate the actual media
objects themselves. For example, a digital photograph simulates an
analog photograph printed on paper; a digital illustration simulates
an illustration drawn on paper; and digital video simulates analog
video recorded on videotape. But that is not how things actually
What software simulates are the physical, mechanical, or
electronic techniques used to navigate, create, edit, and interact
with media data. (And, of course, software also extends and
augments them, as discussed in detail in Part 1.) For example, the
simulation of print includes the techniques for writing and editing
text (copy, cut, paste, insert); the techniques for modifying the
appearance of this text (change fonts or text color) and the layout
of the document (dene margins, insert page numbers, etc.); and
the techniques for viewing the nal document (go to the next page,
view multiple pages, zoom, make bookmark). Similarly, software
simulation of cinema includes all the techniques of cinematography
such as user-dened focus, the grammar of camera movements
9781623568177_txt_print.indd 199 29/04/2013 11:51
(pan, dolly, zoom), the particular lens that denes what part of
a virtual scene the camera will see, etc. The simulation of analog
video includes a set of navigation commands: play forward, play in
reverse, fast forward, loop, etc. In short: to simulate a medium in
software means to simulate its tools and interfaces, rather than its
Before their softwarization, the techniques available in a
particular medium were part of its hardware. This hardware
included instruments for inscribing information on some material,
modifying this information, andif the information was not
directly accessible to human senses such as in the case of sound
recordingpresenting it. Together the material and the instruments
determined what a given medium could do.
For example, the techniques available for writing were deter-
mined by the properties of paper and writing instruments, such as
a fountain pen or a typewriter. (The paper allows making marks
on top of other marks, the marks can be erased if one uses pencil
but not pen, etc.) The techniques of lmmaking were similarly
determined by the properties of lm stock and the recording
instrument (i.e. a lm camera). Because each medium used its own
distinct materials and physical, mechanical, or electronic instru-
ments, each also developed its own set of techniques, with little
Thus, because media techniques were part of specic incom-
patible hardware, their hybridization was prevented. For instance,
you could white out a word while typing on a typewriter and type
overbut you could not do this with the already exposed lm.
Or, you could zoom out while lming progressively revealing more
informationbut you could not do the same while reading a book
(i.e. you could not instantly reformat the book to see a whole
chapter at once.) A printed book interface only allowed you to
access information at a constant level of detailwhatever would
t a two-page spread.
Software simulation liberates media creation and interaction
techniques from their respective hardware. The techniques are
translated into software, i.e. each becomes a separate algorithm.
This was one of the conventions of books which early twentieth-century book
experiments by modernist poets, and designers such as Marinetti, Rozanova,
Kruchenykh, Lissitzky, and others worked against.
9781623568177_txt_print.indd 200 29/04/2013 11:51
And what about the physical materials of diferent mediums? It
may seem that in the process of simulation they are eliminated.
Instead, media algorithms, like all software, work on a single
materialdigital data, i.e. numbers.
However, the reality is more complex and more interesting.
The diferences between materials of distinct media do not simply
disappear into thin air. Instead of a variety of physical materials
computational mediums use diferent ways of coding and storing
informationdiferent data structures. And here comes the crucial
point. In place of a large number of physical materials, software
simulations use a smaller number of data structures.
(A note on my use of the term data structure. In computer
science, a data structure is dened as a particular way of storing
and organizing data in a computer so that it can be used efciently.

The examples of data structures are arrays, lists, and trees. I am
going to appropriate this term and use it somewhat diferently, to
refer to higher-level representations which are central to contem-
porary computational media: a bitmap image, a vector image, a
polygonal 3D model, NURBS models, a text le, HTML, XML
and a few others. Although the IT, media, and culture industries
revolve around these formats, they do not have a standard name
of their own. For me the term representation is too culturally
loaded while the term data type sounds strictly technical. I prefer
data structure because it simultaneously has a specic meaning
in computer science and also a meaning in humanitiesi.e. the
structure part. The term will keep reminding us that what we
experience as media, content or cultural artifact is techni-
cally a set of data organized in a particular way.)
Consider all diferent types of materials that can be used to
create 2D images, from stone, parchment and canvas to all the
dozens types of paper, that one can nd today in an art supply
store. Add to those all of the diferent kinds of photographic lm,
X-Ray lm, lm stocks, celluloid used for animation, etc. Digital
imaging substitutes all these diferent materials by employing just
two data structures. The rst is the bitmapped imagea grid of
discrete picture elements (i.e., pixels) each having its own color
Paul E. Black, ed., entry for data structure in Dictionary of Algorithms and
Data Structures, U.S. National Institute of Standards and Technology, http://xlinux.
9781623568177_txt_print.indd 201 29/04/2013 11:51
or gray-scale value. The second is the vector image, consisting of
lines and shapes dened by mathematical equations.
So what then happens to all the diferent efects that these
physical materials were making possible? Drawing on rough
paper produces diferent efects from drawing on smooth paper;
carving an image on wood is diferent from etching the same
drawing in metal. With softwarization, all these efects are
moved from hardware (physical materials and tools) into
All algorithms for creating and editing continuous-tone images
work on the same data structurea grid of pixels. And while they
use diferent computational steps, the end result of these computa-
tions is always the samea modication in the colors of some of
the pixels. Depending on which pixels are being modied and in
what fashion, the algorithms can visually simulate the efects of
drawing on smooth or rough paper, using oils on canvas, carving
on wood, and making paintings and drawings using a variety of
physical instruments and materials.
If particular medium efects were previously the result of the
interaction between the properties of the tools and the properties
of the material, now they are the result of diferent algorithms
modifying a single data structure. So we can rst apply the
algorithm that acts as a brush on canvas, then an algorithm that
creates an efect of a watercolor brush on rough paper, then a ne
pen on a smooth paper, and so on. In short, the techniques of
previously separate mediums can now be easily combined within
a single image. And since media applications such as Photoshop
ofer dozens of these algorithms (presented to the user as tools
and lters with controls and options), this theoretical possibility
becomes a standard practice. The result is a new hybrid medium
that combines the possibilities of many once-separate mediums.
Instead of numerous separate materials and instruments, we
can now use a single software application whose tools and lters
can simulate diferent media creation and modication techniques.
The efects that previously could not be combined since they were
tied to unique materials are now available from a single pull-down
menu. And when somebody invents a new algorithm, or a new
version of already existing algorithm, it can easily be added to this
menu using the plug-in architecture that became standardized in the
1990s (the term plug-in was coined in 1987 by the developers of
9781623568177_txt_print.indd 202 29/04/2013 11:51
Digital Darkroom, a photo editing application
). And, of course,
numerous other image creation and modication techniques that
did not exist previously can be also added: image arithmetic,
algorithmic texture generation (such as Photoshops Render Clouds
lter), a variety of blur lters, and so on (Photoshops menus
provide many more examples.)
To summarize this analysis: software simulation substitutes
a variety of distinct materials and the tools used to inscribe
information (i.e., make marks) on these materials with a new
hybrid medium dened by a common data structure. Because of
this common structure, multiple techniques that were previously
unique to diferent media can now be used together. At the same
time, new previously non-existent techniques can be added as well,
so long as they can operate on the same data structure.
(Note: many standard contemporary image formats including
the Photoshop .psd format are much more complex than a simple
pixel gridthey can include alpha channels, multiple layers, and
color proles; they can also combine bitmapped and vector repre-
sentations. However, in this discussion, I am only talking about
their common denominatorwhat the algorithms work on when
an image is loaded in memoryan array of pixels holding color
Let us now look at another example of what happens with
physical materials of diferent mediums when they are simulated
in software. Consider 3D modeling software such as Blender,
Maya, 3ds Max Studio, LightWave 3D, or Googles SketchUp.
These applications provide the techniques for dening 3D forms
which were previously hardwired to diferent physical media.
For example, you can use sculpting tools to create a rounded form
as though you are using clay. 3D applications also provide dozens
of new techniques for dening and modifying forms not available
previously: bevel, extrude, spherize, randomize, boolean opera-
tions, smooth, loft, morph, simplify, subdivide, and so on.
As with
image editing software, new techniques can always be added as
long as they operate on the standard data structures already used
by 3D software. (The most common ones are polygonal models
http://en.wikipedia.org/wiki/Digital_Darkroom (February 19, 2012)
A list of a subset of the operations that work on 3D models made from polygons
is provided in http://en.wikipedia.org/wiki/Polygon_modeling#Operations
9781623568177_txt_print.indd 203 29/04/2013 11:51
and NURBS models.
The former consist of at polygons; the
latter are dened by smooth curves.) These data structures are the
new materials that software substitutes for a variety of physical
materials used by humans to create physical 3D forms such as
stone, wood, clay, or concrete.
These two examples of raster image and 3D model data struc-
tures should make clear why it is incorrect to think that computers
always work on a single digital material, so to speak, i.e. the
binary code made from 0 and 1. Of course this is what happens
on a low-level machine levelbut this is largely irrelevant as far
as application software users and people who write this software
are concerned. Contemporary media software contains its own
materials: data structures used to represent still and moving
images, 3D forms, volumes and spaces, texts, sound compositions,
print designs, web pages, and other cultural data. These data
structures do not correspond to physical materials in a 1:1 fashion.
Instead, a number of physical materials are mapped onto a single
structurefor instance, diferent imaging materials such as paper,
canvas, photographic lm, and videotape become a single data
structure (i.e., a bitmapped image). This many to one mapping
from physical materials to data structures is one of the conditions
which enables hybridization of media techniques.
What is a medium?
I have spent considerable time analyzing the specicity of media
software in relation to pre-digital media. This analysis allows us to
better understand why hybridity became the next stage in computer
metamedium evolution and to begin tracing the mechanisms of
hybridization. Let us now see if we can use our new insights to
answer one of the key questions of this book: what is media after
its softwarization?
To avoid confusion: I am not talking about the actual
content of media programs and sources, be they television
programs, newspapers, blogs, or the terrain in Google Earth.
There are already a number of academic disciplines which study
9781623568177_txt_print.indd 204 29/04/2013 11:51
media content and its reception: Media Studies, Communication,
Journalism, Film and TV Studies, Game Studies, Cultural Studies,
and Internet Studies. I am also not going to talk about media
industriesproduction, distribution, reception, markets, economic
aspects, etc., because academic disciplines already analyze these
extensively. However, they usually do not so closely analyze the
tools, technologies and workows used to produce media content.
Even when they do this analysis, it is only done in relation to
the tools of a particular media eld. This is because the modern
academic study of culture follows the commercial culture indus-
tries strict division by type of content. Thus, Game Studies looks
at games, Film and TV Studies looks at lms and television
programs, Design Studies looks at design, Internet Studies looks
at the web, etc. Because of these divisions, these disciplines
ignore the common features of all media and cultural production
being done today which are the result of their reliance on the
same technologyapplication software for media authoring and
editing. (This is one of the reasons why we need a Software Studies
perspectiveto focus our attention on common cultural patterns
related to the use of software technology in all of these diverse
cultural elds and media industries.)
I will also bracket media receptionin other words, I will try to
dene what media is today for its creators as opposed to media
consumers. (While in the 2000s there were many discussions about
the blurring of these denitions because of the falling prices of tools
and emergence of social media, in practice they did not get erased.)
And nally, people who know computer programming and can
create media by writing programs will also have a diferent under-
standing of mediabut the majority of content creators use only
application software.
The universe of the users of application software includes
creative professionals: motion graphics artists, graphic designers,
photographers, video editors, product designers, architects, visual
artists, etc. It also includes prosumers (or pro-ams) making
anime remixes, editing documentary videos which they will upload
to YouTube or Vimeo, shooting photos which they will post to their
Pro Flickr accounts, or uploading their art images to deviantArt.
I want to understand what it means to create media for all
these people as dened by the possibilities of the software they
are usingPhotoshop, Gimp, Illustrator, InDesign, After Efects,
9781623568177_txt_print.indd 205 29/04/2013 11:51
Final Cut, Premiere, CinePaint, Maya, Dreamweaver, WordPress,
Blogger, Flash, OpenOfce, Pages, Microsoft Word, Flame, Maya,
and so on. (And, speaking of Word and other text processing appli-
cations, I should also add millions of people who use it daily and
who, therefore, can be considered experts or at least prosumers in
at least one mediumthat of text authoring and editing.)
Recall one of the dictionary denitions of a medium which
opens Understanding Metamedia (Chapter 2): A specic kind
of artistic technique or means of expression as determined by the
materials used or the creative methods involved. (For example,
the medium of lithography.) Thus, diferent mediums have
diferent techniques, means of expression and creative methods.
These diferences certainly do not disappear when we switch to
software applications. For example, besides the obvious represen-
tational and expressive diference between 3D models and moving
images, a designer who is modeling a game character in Maya
and a designer who is making animation in After Efects will have
access to diferent sets of tools. But are there some conceptual
similarities between the way these two designers will be working
because they both use media software?
In short: what is media today as dened by software applica-
tions used to create and edit it?
As I have already discussed, earlier physical, mechanical, and
electronic media consisted of two components: materials used to
hold information and the tools or equipment used to record, edit,
and view this information. For example, the lm medium used
lm stock for information storage, a lm camera for recording, a
projector for showing lms, and editing devices such as Moviola
and Steenbeck. The medium of hand engraving used metal plates
(typically copper) to hold information, and special, hardened steel
tools to create it by making grooves in the plate.
Do these two components nd their analogs in software?
Here is one answer which we can give to this question: Materials
become data structures; the physical, mechanical, and electronic
tools are transformed into software tools which operate on these
data structures. From this perspective, regardless of the particular
media eld, all designers and artists working with media software
are doing the same thing: using the tools provided by the software
to create, modify, and edit data organized in particular data
9781623568177_txt_print.indd 206 29/04/2013 11:51
This answer is compelling but not precise. As I already discussed,
many materials are mapped into a single data structure. Thus, the
move from physical media to software apps involves a redistri-
bution of the roles previously played by the physical tools and
materials. When I use a watercolor brush and a rough-textured
paper, the resulting brushstrokes are equally the result of the
brush, the liquid, and the paper. But when I use a watercolor
brush in Photoshop, or apply a watercolor lter to an already
existing image, the result is determined solely by an algorithm, that
modies the colors of the pixels in a particular manner. The pixels
are only memory locations, which hold the color valuesthey do
not have any properties of their own, unlike physical materials.
Therefore, we do not have a one-to-one mapping between
physical materials and data structures. The same data structure
(such as a bitmap image) can be used to simulate many imaging
techniques: from watercolor and engraving to photography. To make
this concrete, look at all JPEG images on your computer. Some are
your photographs uploaded from a mobile phone or digital camera;
others are small graphical icons used by various applications; still
others may be the notes you made with a note-taking app, and so
on. The same data structures hold multiple media.
This is why rather than stating that materials turn into data
structures while tools turn into algorithms, it would be more
correct to say that a medium as simulated in software is a combi-
nation of a data structure and set of algorithms. The same data
structure can be shared across multiple medium simulations, but at
least some of the algorithms will be unique to each medium.
We have arrived at a denition of a software medium, which
can be written in this way:
Medium = algorithms + a data structure
Algorithms and data structures happen to be two fundamental
elements of computer programming. In fact, one of the most inu-
ential books in the history of computer science is Niklaus Wirths
Algorithms Plus Data Structure Equals Programs published in
1975. Wirth and other computer scientists conceptualized the
intellectual work of programming as consisting of two inter-
connected parts: creating the data structures which logically t
with the task which needs to be done and are computationally
9781623568177_txt_print.indd 207 29/04/2013 11:51
efcient, and dening the algorithms which operate on these data
We can use this conceptual model of computer programming
to rene our understanding of what media applications do. All
applications, including media software, are computer programs,
so internally they involve algorithms working on data structures.
This point by itself is not very revealing. What is more important is
that these two elements, in my view, also dene a mental model a
user has of the applicationi.e., how the user understands what an
application presents to him/her, and what s/he is doing while using
that application. In other words, the users mental model reects
the abstract structure of a computer program (algorithms operating
on data structure) that drives the particular media application
Inside the application environment, a user is working with
one or more documents which contain content structured in
a particular way. The user is aware of the importance of this
structureeven though media applications do not use the term
data structure. The user understands that the data structure
determines what content can be created and the operations which
can be used to shape and modify it. If I choose vector graphics as
my format, this implies that I will be creating straight and curved
lines and gradients; I will be able to reshape any line in the future
without losing quality; I will also be able to get perfect print
output at any resolution. If I select bitmap image as my format,
I can work with photographs and other continuous-tone images,
blurring and sharpening details, painting over, applying lters and
so on; but the price for this exibility is that the image will exhibit
undesirable artifacts if I enlarge it many times.
(In practice, this
selection is done when a user chooses the primary application to
assemble the project. Choosing Illustrator implies that you will be
working with vector graphics; choosing Photoshop implies that
you will be working with bitmap images. Although each program
also supports working with the opposite image type, the majority
of its functions and its interfaces are organized around its native
An example of a popular vector le format is Illustrators AI format; JPEG and
PSD are examples of bitmap le formats. Certain le formats such as EPS, PDF, and
SWS can hold both vector graphics and bitmap images.
9781623568177_txt_print.indd 208 29/04/2013 11:51
One more reminder because this is important: the term data
structures has a particular meaning in computer science, referring
to how data to be processed by a program is organized. As I have
already said, while I want to retain the core idea of data organi-
zation, I am not interested here in how this actually happens on the
low-level (i.e., whether the program organizes data using arrays,
linked lists, etc.). Instead, I am using this term to refer to the level
of data organization, which is made visible and accessible to a
user and thus becomes a part of his/her mental model of the media
creation and editing process. (For instance, when I work with
bitmap images in Photoshop, I can zoom to examine individual
pixels; I can check resolution of my image in pixels, I can choose
the diameter of my brush, again in pixels; and so on. All this
reminds me that my image is a pixel grid.)
To summarize this discussion: I suggested that both theoretically
and also experientiallyat least for the users who have more than
casual experiences with media applicationsmedia translates
into two parts which work together. One part is a small number of
basic data structures (or formats) which are the foundation of all
modern media software: bitmap image, vector image, 3D polygonal
model, 3D NURBS model, ASCII text, HTML, XML, sound and
video formats, KML, etc. The second part is the algorithms (we can
also call them operations, tools or commands) that operate
on these formats.
The ways in which these two parts are actualized in media appli-
cations require additional discussion. First, diferent applications
often add more details on top on these basic types to give them
additional functionality. For instance, an image as it is dened by
capabilities of Photoshop (a professional and more expensive appli-
cation) is substantially diferent from the way it is dened in the
much less expensive Apples iPhoto or Googles Picasa. As dened
by Photoshop, an image is a complex hierarchical structure. (This
description refers to Adobe Photoshop CS5.5.) At its base is the
basic bitmap image: a grid of pixels. A pixel is a minimal element
that a user can select and modify. This is a basic data structure
Photoshop shares with all other image editors. (Note that while the
user cannot select parts of pixels, the actual algorithms often work
on a sub-pixel level.) A document in Photoshop can contain many
such pixel grids; they are referred to as layers. The layers can be
collected in groups. They can also form layer compsalternative
9781623568177_txt_print.indd 209 29/04/2013 11:51
versions of the composition layout. Any single document layer
can also have many states: a user can make it visible or invisible,
change its transparency, the way it interacts with layers under-
neath, etc. Photoshop also provides special adjustment layers
that do not contain any pixel content; according to the program
documentation, An adjustment layer applies color and tonal
adjustments to your image without permanently changing pixel
Additionally, an image can also have a number of masks
that dene which areas of an image can be edited. Other elements
that Photoshop adds to the basic image structure include paths and
vector-based graphics and types.
Photoshops UI uses a number of windows and menus to
present this complex image structure, with all its possibilities.
The Document window displays the actual composition. The
Layers panel shows all layers, layer efects, and layer groups in the
composition. The Channels panel shows the color components of
an image (such as R, G, B). Each of these windows has a number
of controls. Additionally, various menu items are dedicated to
creating, viewing, and modifying all the possible image parts.
If data structures form one part of a users mental model of
media creation in software applications, the operations that can
be used on these structures comprise the second part. That is,
the user also understands that the process of dening and editing
the content involves the sequential application of diferent opera-
tions provided by an application. Each operation corresponds to
an algorithm which either performs some actions on the already
existing data or generates new data. (Photoshops Wave lter is an
example of the former, while the Render Clouds lter is an example
of the latter.)
In contemporary media software, the tools that are represented
by items in menus are often referred to as commands. The term
tool is reserved for those frequently used operations that are
given their own icons and can be selected directly without having
to navigate through the menus. (I will use the word tool to refer
to both types.) The applications group these related operations
together. For example, Photoshop CS5 collects its key tools in the
Tools Panel; additional tools are found under top-down menus
9781623568177_txt_print.indd 210 29/04/2013 11:51
named Edit, Image, Layer, Select, Filter, and View. Many media
applications also make available additional tools in the form of
scripts that can be run from within the application. For example,
in addition to dozens of commands already available in Photoshop
CS5s menus and panels a user can also run Adobe or third-party
scripts which appear under File > Scripts. These scripts can be
written in JavaScript, VBScript, AppleScript, and other scripting
languages. Finally, people who use command-line interfaces such as
Unix (or Linux) can also use a third type of operationsseparate
software programs which are run directly from a command line.
For example, two very widely used programs for image and video
conversion and editing are ImageMagic and FFmpeg.
Since these
types of programs do not have a GUI interface, they are not suitable
for interactive image editing; however, they excel in automation
and are often used to perform batch operations (such as conversion
from one le format to another) on a large number of les at once.
Regardless of whether the media tools are presented via GUI, as
scripts, or as separate programs available from a command line, they
all have one thing in commonthey can only work on particular
data structures. For instance, image-editing applications dene
dozens of tools for editing bitmap images
but these tools would
not work on vector graphics. In another example, the techniques for
modication of 3D models that dene the volume of an object are
diferent from the techniques that can operate on 3D models that
represent the objects boundary (such as polygonal models).
To make an analogy with language, we can compare data
structures to nouns and algorithms to verbs. To make an analogy
with logic, we can compare them to subjects and predicates. Like
all metaphors, these two highlight and distort, reveal and hide.
However, I hope that they can help me to communicate my key
pointthe dual nature of a medium as dened by software
We have now arrived at one possible answer to the question we
posed in this books introduction: what is media today as dened
by software applications for its creation and editing? As dened
by application software and experienced by users, a medium
http://www.imagemagick.org/, http://ffmpeg.org/
This Wikipedia article lists images editing operations common to these programs:
9781623568177_txt_print.indd 211 29/04/2013 11:51
is a pairing of a particular data structure and the algorithms for
creation, editing and viewing the content stored in that structure.
Now that we have established that computational media
involves combining algorithms and data structures, we can also
better understand the distinction between media-specic and
media-independent techniques that I made earlier. A media-specic
technique is an algorithm that can only operate on a particular data
structure. For example, blur and sharpen lters can only work on
bitmapped images; the extrude operation commonly used in 3D
programs to make 3D models can only be applied to a vector curve.
In contrast, a media-independent technique is a set of algorithms
that all perform a conceptually similar task but are implemented
to work on a number of data structures. I mentioned examples
of these techniques when I introduced this conceptsort, search,
zoom, cut, copy, and paste, randomize, various le manipulations
(copy, email, upload, compress, etc.), etc.
To explain how media-independent techniques can be imple-
mented, let us look at the Copy, Cut and Paste commands. These
operations already existed in some computer text editors in the
1960s. In 19741975 Larry Tesler implemented these commands
in a text editor as part of Xerox PARCs work on a personal
Recognizing that these commands can be used in all
types of applications, the designers of Xerox Star (released in
1981) put dedicated keys for these commands in a special keypad.

The keypad contained keys marked Again, Find, Same, Open,
Delete, Copy, Merge, and Move. A user could select any object
in an application or on the desktop and then select one of these
commands. Xerox PARC team called them universal commands.
Apple similarly made these commands available in all applications
running under its unied GUI but got rid of the dedicated keys.

Instead, the commands were placed under the Edit pull-down menu.
The idea that a user can select objects in any document regardless
of the media, or any le, and use the same set of commands on
http://en.wikipedia.org/wiki/Xerox_Star (February 20, 2012)
For a close-up view of the dedicated keypad for universal commands and demon-
stration of their operations, see the part showing the Xerox Star keyboard in this
video: http://www.youtube.com/watch?v=Cn4vC80Pv6Q&feature=relmfu (August
4, 2012).
9781623568177_txt_print.indd 212 29/04/2013 11:51
these objects is among the most important inventions of the Xerox
PARC team. It gives the user a single mental model of working
with documents across applications, and simplies learning new
This is how the designers of Xerox Star described one of these
universal commands:
MOVE is the most powerful command in the system. It is
used during text editing to rearrange letters in a word, words
in a sentence, sentences in a paragraph, and paragraphs in a
document. It is used during graphics editing to move picture
elements, such as lines and rectangles, around in an illus-
tration. It is used during formula editing to move mathematical
structures, such as summations and integrals, around in an
However, depending on the type of media application and the
kinds of objects a user selects, copy, cut or paste will trigger
diferent algorithms. For example, copying a phrase in a text
document requires diferent sequences of operations than copying
a selection in a bitmap image because the rst is a one-dimensional
sequence of characters, while the second is a set of pixels in a
two-dimensional area. And even within a single application, many
diferent algorithms will be needed to copy diferent kinds of
objects a user can select.
My second example of how implementation of a media-
independent technique involves diferent algorithms that work with
particular media is the generation of random objects. The algorithm
which generates a sequence of random numbers is very simpleit
just calls the random number generator (a function available in all
programming languages) to generate enough numbers, then scales
these numbers within limits specied by the user (for instance, 0 to
1). This part is media-independent. Diferent applications can use
this random number generation function as part of media-specic
algorithms (i.e., algorithms which work on particular data struc-
tures) to create diferent types of content. For example, Photoshop
has a command called Add Noise (located under Filters > Noise)
D. Smith, C. Irby, R. Kimball, B. Verplank, B., E. Harslem, Designing the Star
User Interface, Byte, vol. 7, issue 4 (1982), p. 24282.
9781623568177_txt_print.indd 213 29/04/2013 11:51
A diagram of the Xerox Star UI from D. Smith, C. Irby, R. Kimball, B.
Verplank, B., E. Harslem, Designing the Star User Interface, Byte,
vol. 7, issue 4 (1982), 24282. The universal commands are located in
the dedicated keyword on the left part of the keyboard. (The original
illustration from the article was redrawn in Illustrator.)
9781623568177_txt_print.indd 214 29/04/2013 11:51
which generates a set of random X, Y number pairs. It then uses
them to select specic pixels in the image and convert them to black
or a random primary color (depending on the chosen option). A 3D
modeling application can use the same technique to generate a set
of identical 3D objects randomly located in space. Sound editing
software can generate random sonic noise; and so on.
The implementation of media-independent techniques is struc-
turally similar to various aesthetic systems in art that were
not limited to a particular medium: for instance, baroque,
neo-classicism, constructivism, post-modernism, and remix. Each
system manifested itself across media. Thus, Baroque aesthetics can
be found in architecture, sculpture, painting, and music; construc-
tivism was applied to product design, graphic design, clothing,
theatre, and possibly poetry and lm.
But just as with media-
independent techniques, realizing a particular aesthetic system in
diferent media required some specic artistic devices that explored
possibilities and worked with the limitations of each medium.
File formats
Software uses les to store and transfer data. For example, when
you save a Photoshop image, all its channels, layers, groups, paths,
and other information written to le using a particular format.
Along with data structures, algorithms and UI, a le format is
another fundamental element of computational media. File formats
are the standardized mechanisms for storing and accessing data
organized in a particular structure. Some le formats like .rdf are
in public domain; some like .doc are proprietary. As I will discuss
in more detail in the Design Workow section in Chapter 5,
standardization of le formats is an essential condition for inter-
operability between applications that in turn afects the aesthetics
of media created with these applications. From the point of view of
media and aesthetic theory, le formats constitute the materiality
of computational mediabecause bits organized in these formats is
what gets written to a storage media when a le is saved, and also
Vlada Petric, Constructivism in Film A Cinematic Analysis: The Man with the
Movie Camera (Cambridge: Cambridge University Press, 1993).
9781623568177_txt_print.indd 215 29/04/2013 11:51
because le formats are much more stable than other elements of
computational media (I will explain this below).
Since both the materials and tools of physical media are now
implemented as software, in theory new le formats and new
algorithms can be easily created at any time, and existing ones can
be extended. (Recall my discussion of Permanent Extendibility in
Chapter 1.) However, in contrast to the 1960s and 1970s, when a
few research groups were gradually inventing computational media,
today software is a big global industry. This means that software
innovation is driven by social and economic factors rather than by
theoretical possibilities. As long as le formats are kept constant,
it is easy to add new tools in subsequent software releases and the
old tools will continue to work without modication. Moreover,
in contrast to the period when Kay and others were dening
the rst metamedium, today millions of individual professional
usersas well as design and architecture rms, lm studios, stock
agencies, web design companies, and other creative companies and
groups around the worldstore their work and their assets (3D
models, photographs, print layouts, websites, etc.) as digital les in
particular formats: .doc, .pdf, tif, .html, etc. If le formats were to
change all the time, the value of media assets created or owned by
an individual or a company would be threatened.
As a result, in practice le formats change relatively infrequently.
For example, the JPEG image format has been in use since 1992,
while the TIFF format goes back to 1986. In contrast, the modi-
cation of software tools which can work on lescreating, editing,
displaying, and transmitting themand the creation of new tools
happens at a fast pace. When a company releases a new version
of its application software, it usually adds various new tools and
rewrites some of the existing ones but the le format stays the same.
This stability of media le formats also allows other developers
(both companies and individuals) to create new tools that work on
these formats. In other words, it is one of the conditions that make
constant extendibility of media software possible in practice. Here
is an example of this extendibility: when I visited the plugins area of
Adobes Photoshop website on August 5, 2012 it listed 414 plugin
Given that a typical product can include a dozen or even
ories&marketplaceId=2&offeringtypeid=5 (August 5, 2012).
9781623568177_txt_print.indd 216 29/04/2013 11:51
thousands of lters, presents, or action sets, the total number of
available plugins is likely to run into the hundreds of thousands.
Each le format and its corresponding data structure has
its strengths and weaknesses. A photograph represented as a
bitmapped image can be given a painterly appearance; blurred and
sharpened; composited with another photograph, and so on. All
these operations are much more difcult or even impossible with a
vector image. Conversely, it is much easier to edit complex curves
if they are internally represented by mathematical formulasa
format used by vector drawing programs such as Illustrator and
Inscape. Because many projects call for the combination of efects
only possible with diferent data structures (such as raster and
vector), over time professional software applications were extended
to handle the corresponding le formats in addition to their native
format type. For example, while the majority of Photoshop CS4
tools are geared towards raster images, it also includes some tools
for working with vector drawings. Photoshop can also import
vector graphics, while Illustrator can import bitmap images.
However, this hybridization of software applications does not
change the fact that each separate application tool can only work on
a particular data structure. This is true for universal commands such
as cut, copy, paste, and view, as well as for a multitude of
media-specic commands such as word count, blur, extrude
and echo. Thus, behind both media-independent and media-specic
tools are separate algorithms each designed to work with particular
data structures. However, a user has a diferent understanding of the
two types, since the implementation is not visible directly. The former
bring all media types together conceptually, and even creates an
imaginary horizon where all diferences between them disappear; at
the same time, the latter emphasizes these diferences since they only
become available when a user works with a particular media.
As the name indicates, computer les refer to the paper
les that were the key information management technology of
a mid-twentieth-century ofce when computers were developed.
The word le was used already in 1950 in RCA advertisement
for its new memory vacuum tube; in 1952 the word was used
to refer to information stored on punch cards.
With the devel-
9781623568177_txt_print.indd 217 29/04/2013 11:51
opment of the Web in the 1990s, web documents
such as web
pages became equally important. A web page may consist of a
single HTML le that contains static text, and other media content
stored on a server. Alternatively, a web page may be dynamic,
which means that is constructed when the user accesses its address;
it can also change as a result of user interaction.
Dynamic web
pages can be constricted using client-side scripting (e.g. JavaScript);
they can be also constructed using server-side scripting (PHP, Perl,
Java, and other languages). The two methods can also be combined
using Ajax techniques; for example, the popular web application
that uses Ajax is Google Earth. In 2011, HTML 5, the next gener-
ation of the HTML standard, enabled multimedia and graphics
elements including video, audio, and SVG graphics without the
need for client side plugins. As these and other technologies were
gradually developed and adopted, the identity of the web has been
gradually changingfrom static pages in the rst half of the 1990s
to rich internet applications that match much of the function-
ality of traditional desktop applications.
Given the complexity and variety of web documents and appli-
cations types, the multitude of technologies and techniques for
creating them, and the continuous evolution of both technologies
and web conventions, it would not be appropriate to simply
mechanically map our concept of media data structures to the
web. If we instead focus more on the meaning of data structure
as a mental model of media shared by designers and users, as
opposed to its technical implementations, then this concept does
apply to the web documents and applications. However, rather
than referring to the type of media and its characteristics (text,
bitmap image, vector image, sound, 3D model, etc.) that become
elements of a web document or an application, we can also use
it to describe the interaction possibilities and conventions ofered
by a web document or an application. For example, web pages
typically continue to have hyperlinks that allow a user to go
to related pages. Today a web page may contain social media
buttons which allows users to easily share some content on the
page. Particular genres of web documents and applications ofer
9781623568177_txt_print.indd 218 29/04/2013 11:51
their own interaction possibilities: for example, a blog typically
contains a list of blog posts organized by dates; a webmail appli-
cation contains buttons for responding, forwarding, and archiving
email; and so on.
Although in principle I can go through all the most widely
used types of web documents and applications, and write down
a list of their conventions as they exist right now, this list will
be both very long and no longer accurate by the time my book
manuscript comes out in print. This is one of the reasons why
this book focuses on discussion of media as representation rather
than as communication and interactionbecause the structures
of software-based representations are more stable, less numerous,
less complex, and change much less frequently than software and
network based communication and interaction technologies. This
does not mean that I give up on the project of understanding web
media (including mobile applications, which currently number
in the hundreds of thousands)instead, I hope to do it more
comprehensively in the future, while limiting myself to only brief
discussions in this book.
As I suggested earlier, a users mental model of media creation
and editingboth in the context of a particular application type,
and when dealing with digital media in generalcontains two
fundamental elements, which, both conceptually and practically,
correspond to two elements of computer programming: algorithms
and data structures. When a user selects a particular tool from
a menu and uses it on the part of the document s/he is working
on, the algorithm behind this tool modies the data structure
that stores the content of the part. Most users, of course, do not
know the details of how media software works on a programming
level, and they may be only vaguely aware of what an algorithm
is. (In 2011 I was driving towards San Francisco and saw a big
ad board along the road, which prominently featured the word
algorithmbut I could not stop to snap a picture that I really
wanted to include in this book.) However, unknown to them, the
principles of contemporary computer programming are projected
to the UI levelshaping how users work with media via software
9781623568177_txt_print.indd 219 29/04/2013 11:51
applications practically, and how they understand this process
cognitively. The data structure/algorithm model is one example of
this. We will now look at another example of such projection:
options and their implementation.
One of the principles of modern computer programming
regardless of programming paradigm and programming
languageis the use of parameters.
The popularity of parameters
(which are also called variables or arguments) is due to several
reasons. One is the common modern programming practice of
breaking a program into separate functions. If a program has to
execute the same sequence of steps more than once, a programmer
encapsulates this sequence in a single function which can be
then evoked within the program by its name as often as needed.
(Depending on the programming language, functions maybe be
called procedures, methods, subroutines, or routines.) Dividing
a large program into separate modular functions makes it easier
to write, read, and maintain. (This programming paradigm is
called Procedural Programming.) Functions that perform concep-
tually related tasks (for instance, generating graphics on the
display screen) are collected in software libraries; such libraries
are available for all popular programming languages and their
use greatly speeds up software development. A function denition
typically includes a number of parameters that control the details
of its execution. For example, a function that translates an image
into another format will have a parameter specifying whether the
output format should be JPEG, PNG, TIFF, or another format.
(And if you choose JPEG, you will get another parameter to specify
the level of compression). The second reason for the popularity
of parameters is that many functions (and whole programs) solve
mathematical equations. A formula denes a relationship between
variables. For example, a sine formula looks like this: y=A*sin(w*x
+ O), where A stands for amplitude, w is frequency, and O is phase.
If we implement this formula as a software function, this function
will have parameters for each variable (i.e., w, x and O).
If you do not program, you may still be familiar with the
concept of parameters if you use formulas in Excel or Google
Docs spreadsheet. For example, to generate a column of random
9781623568177_txt_print.indd 220 29/04/2013 11:51
numbers, you can ll its cells with the function RAND(). The
formula does not have any parameters; it simply generates random
numbers that fall between 0 and 1. If you want to generate random
numbers which fall within a diferent range, use a diferent formula
RANDBETWEEN(bottom, top). This formula has two parameters
which specify minimum and maximum values of the numbers of
the range. For example, RANDBETWEEN(10, 20) generates a set
of random values which lie between 10 and 20.
In a modern GUI the parameters, which control program
execution, are called options. Users specify the values for the
options using text boxes, sliders, buttons, drop-down lists, check
boxes, and radio buttons. Sometimes the interface provides a few
possible values and a user can only choose from them; in other
cases she can enter her own value; in still other cases an appli-
cation provides both possibilities. For example, a typical color
picker allows a user to set R, G, and B values to create a unique
color; alternatively s/he can choose from a set of predened color
Use of options greatly expands the functionality of software
applications by allowing the same application to perform a wider
range of actions. For instance, imagine that you need to sort a set
of numbers. Instead of using two separate programs to perform
the sort in ascending and descending order, you can use a single
program and set the parameter value, which would determine what
kind of sort to perform. Or, imagine a round brush tool included
in an image editing applications toolbox. You would not want to
see a separate tool for every possible color, every possible brush
radius, and every possible transparency setting. Instead, a single
tool is given options which control color, radius, and transparency.
What does this mean for media theory? With softwarization,
the implicit possibilities and diferent ways of using physical
tools are made fully explicit. Artistic techniques and means of
expression (see the denition of a medium which opens Part 1) are
given explicit and detailed controls. Like all computer programs
or functions, they now come with many parameters. For instance,
Photoshop CS5s basic brush tool has these parameters: size,
hardness, mode, airbrush capacity, opacity, and ow. Hardness,
opacity, and ow can have any value between 0 and 100; there are
25 modes to choose from; and the diameter can vary from 1 pixel
to 2500 pixels.
9781623568177_txt_print.indd 221 29/04/2013 11:51
Do most users need all these options and such a degree of
precision? Probably not. For example, controlling opacity in 5
percent intervals will probably be quite sufcient for most users.
However, the algorithm that implements opacity behavior is
exactly the same regardless of whether a particular parameter
can have 10 settings, or 100. Since the general logic of software
industry is to always try to ofer users more than the previous
applications version or competitors, it is understandable that the
brush interface gives us the larger number of options and a larger
choice of their values, even though such precision may be not
needed. The same goes for most other tools available in media
software. In this way, the logic of programming is projected to the
GUI level and becomes part of users cognitive model of working
with media inside applications.
While adding more options does require additional programming
labor, ofering more values for these options typically does not. If I
want the brush tool to have a transparency option, I need to write
new code to simulate this behavior. However, giving a user a choice
of 20 or 100 possible values for transparency does not cost me
anything. Like any other media tool a brush tool is an algorithm
that takes some inputs and generates outputs by applying a number
of computations to these inputs. In the case of a brush tool, the
inputs are the values of the options set by the users, and the colors
of the pixels over which the brush moves; the outputs are the new
pixel values. The algorithm does not care what the particular input
values are, and the number of steps required to execute them also
does not change because of particular values.
Pre-industrial physical media tools did not have explicit controls.
There were no numerical parameters to set on a pen, brush, or
chisel. If you wanted to change the diameter of a brush, you picked
a diferent brush. The industrial era introduced new types of media
tools that were mechanical or electronic machines: the electrical
telegraph, photo camera, lm camera, lm projector, gramo-
phone, telephone, television, and video camera. Like all industrial
machines they now came with a few controls. Physically these
controls appeared as knobs, levers, and dials. The next generation
of mediasoftware applicationsgive explicit controls to all the
tools they include. The tools that did not have explicit controls
before now acquired them; the tools that had a few were given
many more. (While machines made from mechanical parts can
9781623568177_txt_print.indd 222 29/04/2013 11:51
have only a limited number of possible settings, software param-
eters typically can have practically unlimited range of values.)
The Language of New Media introduced the idea of
transcodingthe mapping of the conventions and principles
of software engineering to cultural concepts and perceptions.
The explicit parameterization of all media creating and editing
techniques implemented in software is a perfect instance of
transcoding logic. Thus, the fact that Photoshops brush tool
comes with a number of options controls and that opacity and ow
can take any value between 0 and 100 is only partially related to
the meaning of this commanda simulation of diferent physical
brushes. The real reason for this implementation of the command
lies in its identity as a computer program. (As I already discussed,
in modern programming, programs and their parts are given
parameters, which in many cases can take any arbitrary numbers
as inputs.)
Along with being a good example of how principles and
conventions of software development in general are carried over
into media applications and our media lives, parameterization
also exemplies another trend. Seemingly diferent media activ-
itiesediting photographs, creating 3D game characters, editing
a video, or working on a website design or a mobile application
become similar in their logic and workow: select a tool, choose
its parameters, apply it; and repeat this sequence until the project
is done.
Of course, we should not forget that the practices of computer
programming are embedded within the economic and social struc-
tures of the software and consumer electronics industries. These
structures impose their own set of constraints and prerogatives on
the implementation of hardware and software controls, options,
and preferences (all these terms are just diferent manifestations
of software parameters). Looking at the history of media appli-
cations and media electronics devices, we can observe a number
of trends. Firstly, the number of options in media software tools
and devices marketed to professionals gradually increases. For
example, a signicant number of Photoshops tools and lters
have more options and controls than in earlier versions of the
application. Secondly, features originally made available in profes-
sional products later become available in consumer-level products.
However, to preserve the products identities and justify price
9781623568177_txt_print.indd 223 29/04/2013 11:51
diferences between diferent products, a signicant diference in
feature sets is continuously maintained between the two types of
products, with professional software and equipment having more
tools and more parameters than their consumer equivalents. Thus
Photoshop has lots of tools, Photoshop Elements ofers fewer tools
and iPhoto and Picasa have fewer still.
All this is obvioushowever, there is also a third trend, more
interesting for media theory. Following the paradigm already
established by the end of the nineteenth century when the Kodak
company started to market its cameras accompanied by a slogan
you push the button, we do the rest (1892), contemporary
software applications and media devices aimed at consumers
signicantly automate media capture and editing in comparison
to their professional counterparts. For example, during the 2000s
many consumer digital cameras only ofered automatic exposure;
to get full manual controls one had to go to the next price category
of semi-professional cameras. In another example, towards the end
of that decade, consumer cameras started to incorporate automatic
face and smile detectionfeatures that were not available on
expensive professional cameras.
Like any other type of automation, automatic exposure requires
more computing steps than the use of manual settings; the same goes
for applying auto contrast or auto tone commands available
in media software. Thus, if we equate the use of computers with
automation, paradoxically it is the consumers who fully enjoy its
benetsin contrast to professionals who have to labor over all
these manual settings of all these controls but of course, this
is what they paid for: to achieve efects and results that built-in
automation cannot deliver. At the same time, by ofering more
high-level automation in consumer-priced products, the industry
unintentionally undermines the professionals skills. For instance,
today a number of web-based applications and application plug-ins
can take a portrait photo and improve contrast, correct skin tone,
remove skin imperfections and wrinklesall in one step. The
results can be surprisingly goodmaybe not good enough for the
cover of Vogue but quite sufcient for posting the improved photo
to ones prole on a social network site.
Interestingly, in the beginning of the 2010s this trend was partly
reversed. Because of the much larger size of the consumer market
and faster release cycles, hardware and software makers started to
9781623568177_txt_print.indd 224 29/04/2013 11:51
ofer some new features rst in lower-level products and then only
later in more expensive products. For example, Apples Aperture
3 (2010) added options which had already been made available
earlier in a consumer-level iPhoto application in its 2009 release
Faces (face recognition) and Places (a system for identifying geo
locations of the photos and locating them on a map interface).

The metamedium or the monomedium?
Given fundamental similarity in how mediums function as
implemented in software, a logical question arises: do we need
to talk about diferent mediums at all? In other words, is the
computer metamedium a collection of simulated, new, and yet-to-
be-invented media, or is it one mono-medium? Are we dealing
with the metamedium or the monomedium?
We now understand that in software culture, what we identify by
conceptual inertia as properties of diferent mediums are actually
the properties of media softwaretheir interfaces, the tools,
and the techniques they make possible for accessing, navigating,
creating, modifying, publishing, and sharing media documents.
For example, the ability to automatically switch between diferent
views of a document in Acrobat Reader or Microsoft Word is not
a property of electronic documents in general, but as a result of
software techniques whose heritage can be traced to Engelbarts
view control. Similarly, the ability to change the number of
segments that make up a vector curve is not a property of vector
imagesit is an option available in some (but not all) vector
drawing software.
As we have learned, every media software also includes at least
some tools that are not media-specici.e. they are not limited to
working on particular data structures like raster images or vector
drawings. Originally conceptualized at Xerox PARC as ways
to enable users to transfer the cognitive habits learned in using
one application to other applications, today a small number of
Xeroxs universal commands have become a larger number of
http://photo.net/equipment/software/aperture-3/review/ (March 4, 2012).
9781623568177_txt_print.indd 225 29/04/2013 11:51
media-independent media tools and interface techniques which
are shaping how users understand all types of media content.
Despite these fundamental ways in which distinct mediums
become aligned conceptually and practically, I do not want
to abandon the concept of diferent mediums altogether. The
substantial diferences between the operations for media authoring
and editing supported by diferent data structures (in the sense of
this term used here) is one reason to keep this concept. And here
are three additional reasons:
1. Mediums as they are implemented in software are part
of distinct cultural histories that go back for hundreds and often
thousands of years. Electronic text is part of the history of writing;
digital moving images are part of the history of a moving image
which includes shadow plays, phantasmagoria, nineteenth century
optical toys, cinema, and animation; a digital photograph is part of
almost 200 years of photography history. These histories inuence
how we understand and use these media today.
Put diferently, we can say that any lm today exists against the
horizon of all lms ever made and a subset of those lms which a
particular person making or watching a particular lm has seen in
their life; similarly, any digital image exists against the horizon of
all images in a museum without walls (Andr Malraux) of human
visual history. A medium, then, is not just a set of materials and
tools (whether physical, mechanical, electronic, or implemented
in software) and artistic techniques supported by these tools
it is also an imaginary database of all expressive possibilities,
compositions, emotional states and dynamics, representational
and communication techniques, and content actualized in all the
works created with a particular combination of certain materials
and tools.
Systematic digitization of cultural heritage is gradually turning
this imaginary database into a real one. By 2012, Google has
digitized 20 million books, while artstor.org ofered over one
million images of art and architecture digitized by 228 museums
and private collections. However, this process did not start with
digital computers. In the early twentieth century the development
of public art museums, illustrated art magazines and books, lantern
slide lectures, and the academic study of arts in the universities
made large numbers of previously created artworks directly visible
and accessible to the public (as opposed to small numbers only
9781623568177_txt_print.indd 226 29/04/2013 11:51
accessible to art patrons). For instance, American museums started
to develop lantern slide collections of their holdings after 1865;
in 1905 the University of California at Berkeley ofered the rst
architecture history course that used lantern slides.
The digitization of cultural collections since 1990 started to
gradually bring together the materials dispersed among all these
already available resources, making them searchable and accessible
through single websites. For example, by early 2012 Europeana

provided information and links to 20 million digitized cultural
objects including paintings, drawings, maps, books, newspapers,
diaries, music and spoken word from cylinders, tapes, discs
and radio broadcasts, and lms, newsreels and TV broadcasts,
contributed by 1500 European institutions.
In the UK, the BBCs
Your Painting project was set up to ofer free online access to
digital images of all 212,000 paintings from all UK national collec-
tions; this goal was achieved by 2013.
In the US, the Library of
Congress provides access to dozens of digital collections from a
single portal. The collections include 4.7 million high-resolution
newspaper pages (1860) and over one million digital images,
such as 171,000 scanned photo negatives from the Farm Security
Administration/Ofce of War Information program (193545).

These institutional digitized collections are supplemented by
user-uploaded born-digital and digitized cultural artifacts. YouTube
and other video sharing sites contain a substantial sample of all
cinema history in the form of short clips. Flickr has a large number
of photos of artworks taken by museum visitors around the world.
Portfolio sites for professional media creators such as Coroot.com
and Behance.com contain millions of portfolios in art direction,
exhibition design, illustration, interaction design, motion graphics,
and other elds. Manga scanlation websites contain millions of
fan scanned and translated manga pages (as of March 4, 2012,
mangapark.com hosted 5,730,252 pages for 9020 manga series
Scribd.com hosts tens of millions of text documents (you will
http://pro.europeana.eu/web/guest/news/press-releases (March 4, 2012).
http://www.bbc.co.uk/arts/yourpaintings/ (January 31, 2013).
http://www.loc.gov/library/libarch-digital.html (March 4, 2012).
http://www.mangapark.com/ (March 4, 2012).
9781623568177_txt_print.indd 227 29/04/2013 11:51
most likely nd this book there as well); the deviantArt online
community for user-generated art hosts over 100 million submis-
sions from 190 countries (Spring 2012 data).

Although the distribution of what is available in all these online
archives is highly uneven in relation to the types of media, historical
periods, countries, and so on, it is safe to say that today the cultural
classicsi.e. works of famous lm directors, cinematographers,
graphic designers, media designers, composers, painters, writers,
and so onare all available online either in complete form or
in parts. As a result, the idea of a medium in a sense of all
the creative works and possibilities realized so far using a set of
technologies has become quite real. As opposed to holding the
imaginary database of key works in a particular medium in your
head (not the most reliable place for such large-scale storage), you
can now quickly consult the web to locate and study any of them.
Further, you are not limited to the holdings of a single museum or a
library at a time; instead, you can use Europeana, Artstor, or other
large-scale collection aggregators to browse their combined metac-
ollections. For example, while the National Gallery in London
has 2,300 paintings, Your Painting site from BBC ofers images of
212,000 paintings from all UK national collections.
2. We can also use the term medium to refer to a presentation/
interaction platform. If you think of iOS and Android platforms
(each containing mobile devices, an operating system, and apps)
as examples, you are rightbut I also would like to use the word
platform in a more general sense. Medium as a platform refers
to a set of resources that allows users to access and manipulate
content in particular ways. Understood in that way, a white cube in
a modern art gallery is a medium; so are a modern movie theatre,
a print magazine, network television, and a DVD. (Note that just
as iOS and Android are ecosystems that combine many elements,
many older media platform work similarly. Movie theatres, lm
production, distribution, and publicity form the cinema platform;
so do television production, distribution, and TV sets.)
This meaning of medium is related to the twentieth-century
concept of media in Communication Studies, which came
from Shannons Information Theory: the storage and transmission
9781623568177_txt_print.indd 228 29/04/2013 11:51
channels and tools used to store and deliver information. However,
if the latter concept includes both storage and transmission
channels and tools, my presentation platform puts more focus
on the reception technologies. These technologies may include
special spaces, architecture (for example, large media surfaces
which are increasingly becoming part of buildings interiors and
exteriors), sensors, lights, and, of course, media viewing/editing/
sharing devices and apps. Presentation platforms also program
particular patterns of behavior: one walks through and touches
architecture; stays silent in a movie theatre; interacts with family
members while watching TV, moves ones body in front of an inter-
active wall; etc.
During the nineteenth and twentieth centuries, presentation
platforms were closely tied up with particular types of media
content. Art museums displayed paintings and sculptures (and later
performances and installations); newspapers published texts and
images; TV presented television shows, news programs, and lms.
The gradual addition of media viewing and playing capabilities to
computers (and later laptops, mobile phones, tablets, media players
and other computer-based devices) loosened this connection.
Distribution, storage, and presentation of diferent media types were
no longer tied up to particular technologies and particular presen-
tation platform. (This separation parallels the similar process of
softwarization I have discussed in detail: the tools and techniques
for media authoring and editing have become liberated from their
reliance on particular physical and electronic technologies.)
Today, we can access most types of media on every computer
platform. You can view images, videos, text documents and
maps inside email, in a browser, on your notebook, a PC, laptop,
tablet, mobile phone, internet-enabled TV, or an in-car or in-ight
entertainment system. What diferentiates these devices is neither
the types of contents they can play, nor the basic interfaces they
provide for viewing and interacting with content, but rather the
relative ease with which one can navigate various media.
For instance, my 2011 Samsung LCD-TV came with a full web
browserhowever it was much better at playing cable TV and
Netix content than web browsing experience, which was quite
hard with a TV remote control. In another example, the relatively
small screen of a mobile phone in the early 2010s, its less powerful
processor, and its smaller amount of RAM makes it a less-than-ideal
9781623568177_txt_print.indd 229 29/04/2013 11:51
platform for editing feature lms or doing CAD (Computer Aided
Design). At the same time, the same small size gives a mobile
phone platform many advantages. In many countries it is socially
acceptable for a person to engage in chat or send and receive SMS
on their phone during a social mealbut doing it on a laptop would
be inappropriate. The size of a mobile phone also makes it a perfect
device for location-based social networking (e.g., Foursquare) and
other services: recommending social events in a city, following
friends on a map, playing location-aware games, and so on.

Similar to mobile phones, many platforms take advantage of
their social settings by adding unique features. For example, some
airline in-ight entertainment electronics systems (e.g., Virgin
Americas and V Australias RED Entertainment System
) allow
passengers to chat with other passengers or participate in multi-
player games with them.
As a media presentation and interaction platform, each type
of computer-based consumer device has its diferences. Currently
(early 2013), some mobile platforms such as iOS do not allow users
to save documents directly to the device; instead, les can only be
saved inside their respective apps.
Some media devices such as
e-book readers, video players, audio players, digital billboards, and
game consoles often can only play a few (or even just one) type of
media content. (While consumer electronics companies are engaged
in the ongoing convergence war, gradually adding ability to play
all media types to most devices, this trend does not afect every
single device.)
Given the diferences in their physical appearance (size, weight,
form factor), physical interface (touchscreen, keyboard, remote
control, voice input, motion sensing) and media playback/editing/
sharing/networking capabilities, it is tempting to think of each
type of device as a diferent medium. As presentation/inter-
action platforms these devices provide distinct user experiences
and encourage distinct sets of media behaviors (sharing location,
working, chatting, etc.). However, we should also remember that
since they all use the same technologiescomputers, software, and
networksthey also share many fundamental features.
9781623568177_txt_print.indd 230 29/04/2013 11:51
Given this, we may want to think of all these presentation/inter-
action platforms using the idea of family resemblance articulated
by many nineteenth-century thinkers and later by Wittgenstein
that things may be connected by a series of overlapping similarities,
where no one feature is common to all.
But this would not be
accurate. Prototype theory, developed by the 1970s by psychologist
Eleanor Rosch and other researchers, may fare better. Based on
psychological experiments, Rosch showed that for a human mind,
some members of many semantic categories are better representa-
tives of these categories than other members (or example, a chair is
more prototypical of the category furniture than a mirror.
If we consider the best current implementation of Kays
vision augmented with networking and sharing capabilities as the
prototype (i.e. the most central member of the category) of the
computer metamedium (for me, my current Apple laptop would
qualify as such a prototype, but any full-featured laptop will
also do), then all other computer devices can be situated at some
distance from the prototype based on how well they instantiate
this vision. Of course, it is also possible to argue that none of
the current computers or computer devices realizes Kays vision
sufciently, because casual users cannot easily program them and
invent new media. (Appropriately, Kay named his 1997 Turing
Award lecture The Computer Revolution Hasnt Happened
) In this interpretation, Kays Dynabook is the imaginary
ideal prototype, and each realized computer device is situated at
some distance from it.
3. Another important meaning of the concept medium is
related to human sensory systems, which acquire and process infor-
mation in diferent ways. Each sensory system contains sensory
receptors, neural pathways, and particular parts of the brain
responsible for further processing. Traditional human cultures
recognized ve senses: sight, hearing, taste, smell, and touch.
Additionally, humans can also sense temperature, pain, positions
of body parts, balance, and acceleration.
Eleanor Rosch, Cognitive Representation of Semantic Categories. Journal of
Experimental Psychology: General 104, no.3, (September 1975): pp. 192233.
(March 5, 2012).
9781623568177_txt_print.indd 231 29/04/2013 11:51
Because the concept of senses has been important for ancient and
modern Western philosophy, Buddhist thought, and other intel-
lectual traditions, the discussions of senses in relation to art and
aesthetics have also been very extensive, so to enter them seriously
would require its own book. Instead, I will limit myself to a short
discussion of a more recent research in cognitive psychology that
can be used to support the idea of multiple mediums: how diferent
types of information are represented and manipulated in the brain.
The important question debated in cognitive psychology for
forty years is whether human cognition operates on more than
one type of mental representation. One view is that the brain only
uses a single propositional representation for diferent types of
information. According to this view, what we experience as mental
images are internally translated by the mind into language-like
The alternative view is that the brain represents
and processes images using a separate representational system.
After decades of psychological and neurological studies, the current
consensus supports this second view.
Thus, it is believed the brain
operates on and maintains mental images as mental image-like
wholes, as opposed to translating them into propositions.
If we accept this view that language, and images/spatial forms
require the use of diferent mental processes and representa-
tions for their processingpropositional (i.e., concept-based, or
linguistic) for the former, and visual/spatial (or pictorial) for the
latter, it helps us to understand why the human species needed
both writing and visual/spatial media. Diferent mediums allow us
to use these diferent mental processes. In other words, if media
are tools for thought (to quote again the title of a 1984 book
by Howard Rheingold about computers
) through which we
think and communicate the results of our thinking to others, it is
logical that we would want to use the tools to let us think verbally,
visually, and spatially.
Tim Rohrer, The Body in Space: Dimensions of Embodiment, in Body,
Language and Mind, vol. 2, eds. J. Zlatev, T. Ziemke, R. Frank, R. Dirven (Berlin:
Mouton de Gruyter, 2007).
Howard Rheingold, Tools for Thought: The History and Future of
Mind-Expanding Technology, a revised edition of the original book published in
1985 (The MIT Press, 2000).
9781623568177_txt_print.indd 232 29/04/2013 11:51
An hypothesis about the existence of propositional and pictorial
representations in the mind is one example of a number of
theories which all share the basic belief that human thinking
and understanding are not limited to the use of language. For
example, in 1983 Howard Gardner proposed a theory of multiple
intelligences, which included eight diferent categories: bodily-
kinesthetic, verbal-linguistic, visual-spatial, musical, interpersonal,
logical-mathematical, naturalistic, and intrapersonal. Recall also
that Alan Kay based the design of the GUI on the work of the
psychologist Jerome Bruner who postulated the existence of three
modes of representation and cognition: enactive (action-based),
iconic (image-based) and symbolic (language). And while in Xerox
PARCs implementation the appeal to the rst modality was limited
to the selection of objects on the screen using a mouse, more
recently they were joined by touch interfaces and gesture interfaces.
Thus, the interfaces of modern computers and computer-based
devices are themselves prime examples of a multiple media at
workadding over time more media as interaction mechanisms,
rather than converging to a single one, like written language
(original UNIX and other OS from 1960s1970s) or speech (Hal
from Kubricks 2001).
The evolution of media species
As we see, while softwarization redenes what mediums are and
how they interact, it does not erase the idea of multiple distinct
mediums. In fact, in contrast to the idea of convergence popular
in the 2000s in discussing the coming together of computers,
television, and telephony, I would like you to think of the compu-
tational media using the concept of biological evolution (which
implies increasing diversity over time). And here comes the ultimate
difculty with continuing to use the term medium as a useful
descriptor for a set of cultural and artistic activities. The problem
is not that multiple mediums converge into one monomedium
they do not. The problem is exactly the opposite: they multiply
to such extent that the term loses its usefulness. Most large art
museums and art schools usually have between four and six
departments which supposedly correspond to diferent mediums
9781623568177_txt_print.indd 233 29/04/2013 11:51
(for example, San Franciscos Museum of Modern Art divides its
collection into painting and sculpture, photography, architecture
and design, and media arts
)and this is OK. We can still use
unique names for diferent mediums if we increase their number
to a couple of dozens. But what to do if the number goes into
thousands and tens of thousands?
And yet this is exactly the situation we are in today because of
softwarization. Extrapolating from the dictionary denition of a
medium that opens Part 1, we can say that diferent mediums have
sufciently dissimilar representational, expressive, interactive and/
or communicative capacities. (How much of a diference is needed
for us to declare that we have two mediums rather than one? This
is another fundamental question that makes it challenging to use
the term medium in software culture.) Consider all the factors
involved in gradual and systematic expansion of these capacities:
permanent extendibility of software; the development of new
types of computer-based and network enabled media devices (game
platforms, mobile phones, cameras, e-book readers, media players,
GPS units, digital frames, etc.) and the processes of media hybridi-
zation manifested in software applications, technology prototypes,
commercial and artistic projects Do we get a new medium every
time a new representational, expressive, interaction or communi-
cation functionality is added, or is a new combination of already
existing functions created? For example, does the addition of
voice interface to a mobile device create a new medium? What
about an innovative combination of diferent visual techniques
in a particular music video? Does this music video dene a new
medium? Or what about the versions of Google Earth that run on
a computer and support diferent types of layers, and the iPhone
version which in 2008 only supported Wikipedia and Panoramio
layers? Are these two diferent mediums?
If we continue to hold to our extended medium denition
(extrapolated from the dictionary denition), we would have to
answer yes to all these questions. Clearly, we do not want to do this.
Which means that the conceptual foundation of media discourse
(and media studies)the idea that we can name a relatively small
http://www.sfmoma.org/explore/collection (March 6, 2012).
9781623568177_txt_print.indd 234 29/04/2013 11:51
number of distinct mediumsdoes not hold anymore. We need
something else instead.
To explain this in an alternative way, when we considered only
one aspect of media ecosystemmedia designwe were able
to adopt the traditional understanding of medium (materials
+ tools) to describe the operations and interfaces of application
software. We did this by proposing this new denition: medium =
algorithms + a data structure. Following this perspective, we can
refer to simple text, formatted text, vector graphics, bitmap images,
polygonal 3D models, spline-based models, voxel models, wave
audio les, MIDI les, etc. as separate mediums. But even with this
approach it was already difcult to cover the design of interactive
applications and websites. When we start considering the larger
ecosystem of the proliferating devices, network services, interface
technologies, media projects, and over one million apps for mobile
platforms available to consumers, the concept can no longer be
stretched to describe them in a meaningful way. In this section, I
will explore one way to think beyond medium.
As I suggested above, the model of multiple species related via
evolutionary development that we can borrow from biology ofers
a plausible alternative. This model is useful for thinking about both
media authoring/editing applications and particular media projects/
products (which, after all, are also software applications but more
specializedif media applications are content agnostic, projects
typically ofer particular content). The key advantages of a species
model over a mediums model are their large numbers (Earth
contains many million speciesat least for now); their genetic
links (which implies signicant overlap in features between related
species); and the concept of evolution (which implies constant
development over time and gradually increasing diversity).
Each of these advantages is equally important. Instead of trying
to divide the extremely diverse media products of software culture
into a small number of categories (e.g. mediums), we can think
of each distinct combination of a particular subset of all available
techniques as a unique media species. Of course, a software
application or a project/product is not limited to remixing already
existing techniques; it can also introduce new one(s), which may
then reappear in other applications/products. Such new intro-
ductions can perhaps be thought as new genes (keeping in mind
all the limitations of our use of these biological metaphors). If
9781623568177_txt_print.indd 235 29/04/2013 11:51
contemporary sciences such as evolutionary biology, genetics and
neuroscience can describe and map (or at least, work towards
describing and mapping) millions of distinct species, 3 billions of
DNA pairs in the human genome, and 100 billion neurons in a
cerebral cortex, why cannot media theory and software studies deal
with the diversity and variability of software culture by providing
richer classications than the kinds we currently use? (For inspi-
rational examples from live sciences, see the Human Connectome
Project to create a comprehensive map of the adult brain,
even more ambitious, The Blue Brain Project to create a completely
realistic software simulation of the brain, accurate on a molecular
Equally valuable is the notion that related species share many
features. Eighteenth- to twentieth-century aesthetic tradition, from
Gotthold Lessing to Clement Greenberg, repeatedly insisted on
opposing a small number of mediums to each other, thus seeing
them as distinct and non-overlapping categories. This trend inten-
sied in European modernism of the 1910s1920s, when artists
tried to reduce every medium to its unique qualities. To do so, they
gave up representation and concentrated on the material elements
thought to be unique to each medium. Poets, such as Russian
futurists, were experimenting with sounds; lmmakers proposed
that the essence of cinema was movement and temporal rhythm
(French lm theory of the 1920s
) or montage (Kuleshovs group
in Russia); and painters were exploring pure colors and geometric
forms. For example, one of the pioneers of abstract art, painter
Wassily Kandinsky, published two articles in 1919 with the titles
which clearly signal his program to articulate an artistic language
consisting only of basic geometric elements: Small Articles About
Big Questions. I. About Point, and II. About Line.
Although artists in the second half of the twentieth century
systematically revolted against this trend, instead embracing
http://www.humanconnectomeproject.org/about/ (March 7, 2012).
http://bluebrain.epfl.ch/ (March 7, 2012).
See Jean Epstein, On Certain Characteristics of Photognie, in French Film
Theory and Criticism, ed. Richard Abel (Princeton: University of Princeton
Press, 1988), 1: pp. 31418; Germaine Dulac, Aesthetics, Obstacles, Integral
Cingraphie, in French Film Theory and Criticism, 1: 38997.
Kandinsky further developed the ideas in these articles in his 1926 book Point
and Line to Plane.
9781623568177_txt_print.indd 236 29/04/2013 11:51
mixed media, assemblage, multimedia art, and installa-
tions, this did not lead to a new, richer system for understanding
art. Instead, only a few new categories were created (e.g., the ones
I just listed), and all the extremely diverse new art objects created
by these artists were placed in these categories.
Evolutionary biology instead gives us a model of the much
larger space of objects, which overlap in their identities. This model
ts much better with my theory of software culture as a large
and continuously growing pool of techniques that can enter into
numerous combinations, in the form of applications, and projects/
products created with them, or through custom programming.
Obviously, the majority of these media species share at least
some techniques. For example, all applications with a GUI interface
designed to run on a full-size computer screen will use similar
interaction techniques. To take another set of media species I have
already analyzed as examples of media hybridity, Fujihatas Alsace,
Sauters Invisible Shape, and Google Earth all share a technique of
embedding videos inside a virtual navigable space. Both Fujihata
and Sauter construct unique interfaces to video objects in their 3D
worlds. In contrast, Google Earth inhabits the YouTube video
interface when you embed video from YouTube in a placemark
(although you have some control over the embedded player charac-
). This ability to embed the YouTube video player ofers yet
another example of a technique now used in numerous websites
and blogs; so is the use of APIs ofered by most major social media
services and media sharing sites.
Finally, another attraction of the species model is the idea
of evolutionnot the actual mechanisms of biological evolution
on Earth as theorized and argued over by scientists (because these
mechanisms do not t technological and cultural evolution), but
rather the image of gradual but continuous temporal development
and increased variability and speciation (emergence of new species)
this idea implieswithout implying progress. Without subscribing
to the theory of memes, we can name a number of ways in which
new techniques are transmitted in software culture: new projects
and products that are seen by other designers and programmers;
scientic papers in computer science, information science, HCI,
YouTube Help, Embedded player style guide, http://support.google.com/
youtube/bin/answer.py?hl=en&answer=178264 (March 7, 2012).
9781623568177_txt_print.indd 237 29/04/2013 11:51
media computing, visualization and other related elds; and also
the code itself. This last mechanism is particularly important for
us because, in contrast to previous media, it is unique to software
culture (at the same time, it has a direct parallel with genetic
code). As new techniques for media editing, analysis, interaction,
transmission, retrieval, visualization, etc. become popular, they are
coded in multiple programming languages and scripting environ-
ments (Java, C++, JavaScript, Python, Matlab, Processing, etc.) and
become available as either commercial or, increasingly, open source
software libraries ready for download online. If a programmer
working on a new application, website, or any other media project
wants to use any of these techniques, s/he can include the necessary
functions in his/her own code.
Traditionally, new cultural techniques were transmitted via
imitation and training. Professional artists or artisans saw something
new and they imitated it; pupils, assistants, or students would learn
techniques from their teachers; art students would spend years
making copies of famous works. In either case, the techniques
moved from mind to mind and from hand to hand. Modern media
technologies add a new mechanism for cultural transmissionthe
interfaces and manuals of media devices, which carry with them
the suggested, proper, ways of using these devices. All these mecha-
nisms are at work today as well (of course, speeded up by the web).
However, they are also joined by a new onethe transmission of
cultural techniques via algorithms and software libraries.
This does
not necessarily mean that such transmission does not introduce
changesprogrammers can always modify the existing code to the
needs of their project. More important I think is the fundamental
modularity of cultural objects created via this mechanism. On the
level of techniques, a cultural object becomes an agglomeration
of functions drawn from software librariesthe DNA it shares
with many other objects that use the same techniques. (While
commercial media products also widely use content elements such
as photos purchased from stock libraries, this kind of modularity
Jeremy Douglass suggested that we can study the propagation of techniques
across software culture by tracking the use of particular software libraries and
their functions across programs. Jeremy Douglass, presentation at SoftWhere
2008 workshop, University of California, San Diego, May 2008, http://workshop.
9781623568177_txt_print.indd 238 29/04/2013 11:51
is not openly acknowledged.) Because techniques are coded as
software functions, this cultural modularity is closely linked to the
principle of modularity in modern computer programmingthe
concept that a program should contain a number of self-contained
parts. If a media project or an app introduces a new technique (or
techniques) and they appear to be valuable, often the programmers
make them available as stand-alone functions which then enter
the overall pool of all techniques available within the computer
metamediumand thus easing the ways for others to adopt this
technique in creating new projects.
I would like to end this chapter by quoting the historian Louis
Menand who explains that while prior to Darwin scientists viewed
species as ideal types, Darwin shifted the focus to variationwhich
for me is an overarching reason for thinking about media in the
software age through the terms of evolutionary biology:
Once our attention is redirected to the individual, we need
another way of making generalizations. We are no longer
interested in the conformity of an individual to an ideal type;
we are now interested in the relation of an individual to the
other individuals with which it interacts Relations will be
more important than categories; functions, which are variable,
will be more important than purposes; transitions will be more
important than boundaries; sequences will be more important
than hierarchies.

Louis Menand. The Metaphysical Club (New York: Farrar, Straus, and Giroux,
2001), p. 123.
9781623568177_txt_print.indd 239 29/04/2013 11:51
9781623568177_txt_print.indd 240 29/04/2013 11:51
Software in action
9781623568177_txt_print.indd 241 29/04/2013 11:51
9781623568177_txt_print.indd 242 29/04/2013 11:51
Media design
We shape our tools and thereafter our tools shape us.
Marshall MLuhan, Understanding Media (1964)
After Eects and the invisible revolution
Media hybrids are not limited to particular software applica-
tions, user interfaces, artistic projects, or websites. If I am right
in suggesting that hybridity represents the next logical stage in
the development of computational media, following the rst
stage of simulating individual physical media in a computer,
then we can expect to nd it in many cultural areas. And
this is indeed the case. In this chapter I will look at a single
cultural area in depthmoving image designanalyzing how
the creation and aesthetics of moving images changed dramati-
cally in the 1990s.
Around the middle of the 1990s, the simulated physical
media for moving and still image production (cinematography,
animation, graphic design, typography), new computer media (3D
animation), and new computer techniques (compositing, multiple
levels of transparency) met within a single software environment
compatible software programs running on a personal workstation
or a personal computer. Filmmakers, animators, and designers
started to systematically work in this environment, using software
both to generate individual elements and to assemble all elements
9781623568177_txt_print.indd 243 29/04/2013 11:51
together. The result was the emergence of a new visual language
that quickly became the norm.
Today this language dominates the visual media produced in
dozens of countries. We see it daily in commercials, music videos,
TV graphics, lm titles, interactive interfaces of mobile phone
and other devices, dynamic menus, animated web pages, graphics
for mobile media content, and other types of animated, short
non-narrative lms and moving-image sequences being produced
around the world by media professionals including companies,
individual designers and artists, and students. All in all, I estimate
that at least 50 percent of short moving image works follow this
language. In this chapter I will analyze what I perceive to be some
of its dening features: new hybrid visual aesthetics; systematic
integration of previously non-compatible media techniques; use
of 3D space as a platform for media design; constant change on
every visual dimension; and amplication of cinematographic
The new hybrid aesthetics exist in endless variations but its
basic principle is the same: juxtaposing previously distinct visual
aesthetics of diferent media within the same image. This is
an example of how the logic of media hybridity restructures a
large part of culture as a whole. The languages of design, typog-
raphy, animation, painting, and cinematography meet within
the computer. Therefore, along with being a metamedium as
formulated by Kay, we can also call a computer a metalanguage
platform: the place where many cultural languages of the modern
period come together and begin creating new hybrids.
How did this language come about? I believe that looking at
the software involved in the production of moving images goes
a long way towards explaining why they now look the way they
do. Without such analysis we will never be able to move beyond
the commonplace generalities about contemporary culturepost-
modern, global, remix, etc.to actually describe the particular
languages of diferent design areas, to understand the causes behind
them and their evolution over time. (In other words, I think that
software theory, which this book aims to theorize and put in
practice, is not a luxury but a necessity.)
Although the transformations I will be discussing involved
many technological and social developmentshardware, software,
production practices, and workows, new job titles and new
9781623568177_txt_print.indd 244 29/04/2013 11:51
Opening titles animation for television series Mad Men. Imaginary
Forces, 2007.
9781623568177_txt_print.indd 245 29/04/2013 11:51
professional eldsit is appropriate to highlight one particular
software application which was at the center of the events. This
application is After Efects. In this chapter we will take a close
look at its interface, the tools, and its typical use in media design.
Introduced in 1993, After Efects was the rst software application
designed to do animation, compositing, and special efects on the
personal computer.
Its broad efect on moving image production
can be compared to the efects of Photoshop and Illustrator on
photography, illustration, and graphic design.
After Efects certainly has its competitors. In the 1990s,
companies also widely used more expensive high-end software
such as Flame, Inferno, or Paintbox that run on specialized
graphics workstations, and they are still utilized today. In the
2000s, other programs in the same price category as After Efects
such as Apples Motion, Autodesks Combustion, and Adobes
Flash have also challenged After Efects dominance. However,
because of its afordability and length of time on the market After
Efects continues to be the most popular, most widely used, and
best-known application. Consequently, After Efects will be given
a privileged role in my account as both the symbol and the key
material foundation that made the wide-reaching transformation in
moving image culture possible. (When I searched for best motion
graphics software on the web and checked the answers to this
question on a variety of forums, the rst program mentioned was
always After Efects. As one person put it on one of these forums,
Its pretty much the gold standard. Learn it, love it.
In another
example, when in 2012 Imaginary Forcesthe company most
closely associated with the rise of motion graphics in the 1990s
posted descriptions of new jobs in its Los Angeles and NYC ofces
The NewTek Video Toaster released in 1990 was the rst PC-based video
production system that included a video switcher, character generation, image
manipulation, and animation. Because of their low costs, Video Toaster systems
were extremely popular in the 1990s. In the context of this book, After Efects is
more important because, as I will explain below, it introduced a new paradigm for
moving image design that was diferent from the familiar video editing paradigm
supported by systems such as Toaster and Avid.
John Waskey, http://www.quora.com/What-is-the-best-software-for-creating-
motion-graphics (March 4, 2001).
9781623568177_txt_print.indd 246 29/04/2013 11:51
for designers and animators, it listed only one required software
application for 2D moving image production: After Efects.
As I will show, After Efects UI and tools bring together
fundamental techniques, working methods, and assumptions of
previously separate elds of lmmaking, animation and graphic
design. This hybrid production environment, encapsulated in a
single software application, is directly reected in the new visual
language it enablesspecically, its focus on exploring aesthetic,
narrative, and afective possibilities of hybridization.
The shift to software-based tools in the 1990s afected not only
moving image culture but also all other areas of media design. All
of them adopted the same type of production workow. (When
the project requires many people and many media elements, the
production workow is called a pipeline.) In this workow,
designers typically either combine elements created in diferent
software applications, or move the whole project from one appli-
cation to the next to take advantage of their unique possibilities.
And while each design eld also employs its own specialized appli-
cations (for instance, web designers use Dreamweaver, architects
use Revit, and visual efects artists use Nuke and Fusion), they also
all use a number of common applications: Photoshop, Illustrator,
Final Cut, After Efects, Maya, 3ds Max, and a few others. (If
you use open source software like Gimp and CinePaint instead of
these commercial applications, your list of key applications will be
diferent, but the workow would not change.)
The adoption of a production environment and workow that
uses a small number of compatible applications in all areas of
creative industries has had many fundamental efects. The profes-
sional boundaries between diferent design elds have become less
important. A single designer or a small studio may work on a music
video today, a product design tomorrow, an architectural project
or a website design the day after, and so on. Another previously
fundamental distinctionthe scale of a projectalso now matters
less, and sometimes not at all. Today we can expect to nd exactly
the same shapes and forms in very small objects (like jewelry),
small and medium sized objects (tableware, furniture), large
buildings, and even urban designs. (Lifestyle objects, furniture, and
http://www.imaginaryforces.com/jobs/los-angeles/designer/, http://www.imaginary
forces.com/jobs/new-york/2d-animator/ (October 31, 2012).
9781623568177_txt_print.indd 247 29/04/2013 11:51
architectural and urban design by Zaha Hadids ofce illustrate this
A comprehensive discussion of these and many other efects of
software adoption would take more than one book, and therefore
in this chapter I only focus on the impact of the software-based
workow on contemporary media design. As we will see, this
workow shapes contemporary design in a number of ways. On
the one hand, never before in the history of human visual commu-
nication have we witnessed such a variety of visual forms as today.
On the other hand, exactly the same techniques, compositions
and iconography can now appear in any media. To evoke the
metaphor of biological evolution again, we can say that despite
seemingly innite diversity of contemporary media, visual, and
spatial species, they all share some common DNA. Many of the
species also share a basic design principle: integration of previously
non-compatible techniques of media designa process which I am
going to call deep remixability. Thus, a consideration of media
authoring software and its usage in production would allow us to
begin constructing a map of our current media/design universe,
seeing how its species are related to each other and revealing the
mechanisms behind their evolution.
The adoption of After Efects and related software in the
second part of the 1990s quickly led to the adoption of a special
term to designate new animated visuals motion graphics.
Concisely dened in 2003 by Matt Frantz in his Masters thesis as
designed non-narrative, non-gurative based visuals that change
over time,
motion graphics include lm and television titles, TV
graphics, dynamic menus, graphics for mobile media content, and
other animated sequences. Typically motion graphics appear as
parts of longer pieces: commercials, music videos, training videos,
narrative and documentary lms, interactive projects. Or at least,
this is how it was in 1993; since that time the boundary between
motion graphics and everything else has progressively become
harder to dene. Thus, in the 2008 version of the Wikipedia article
about motion graphics, the authors wrote that the term motion
graphics has the potential for less ambiguity than the use of
Matt Frantz Changing Over Time: The Future of Motion Graphics, MFA
Thesis, 2003, http://www.mattfrantz.com/thesisandresearch/motiongraphics.html
9781623568177_txt_print.indd 248 29/04/2013 11:51
the term lm to describe moving pictures in the 21st century.

Certainly, today numerous short moving image works combine
live footage, 2D animation, 3D animation, and other techniques
equally (as opposed to privileging live action cinematography as
many feature lms still do), so they all can be called motion
Why did I select motion graphics as my central case study of
this book, as opposed to any other area of contemporary culture
similarly afected by either the switch to a software-based production
process, or native to computers? The examples of the former area
sometimes called going digital are architecture, graphic design,
product design, information design, and music; the examples of
the latter area are game design, interaction design, user experience
design, user interface design, web design, and interactive information
visualization. Obviously, most of the new design areas which have
interaction or information as part of their titlesand which
emerged since middle of the 1990s have been equally ignored by
cultural critics, and thereforedemand as much attention.
My reason has to do with the diversity of new formsvisual,
spatial, and temporalthat developed during the rapid growth
of the motion graphics eld after the introduction of After Efects.
If we approach motion graphics in terms of these forms and
techniques (rather than only their content), we will realize that they
represent a very signicant turning point in the history of human
communication. Maps, pictograms, hieroglyphs, ideographs, various
scripts, alphabet, graphs, projection systems, information graphics,
photography, modern language of abstract forms (developed rst
in European painting in the 1910s and subsequently adopted in
graphic design, product design and architecture in the 1920s), the
techniques of twentieth-century cinematography, 3D computer
graphics, and of course, a variety of born digital visual efects
in short practically all communication techniques developed by
humans until the 1990sare now routinely combined in motion
graphics projects. Thus, almost all of the previously separate
semiotic resources become options within the users palette (or
toolbox, to use the standard metaphor deployed in media devel-
opment software). Linguistic, kinetic, spatial, iconic, diagrammatic,
9781623568177_txt_print.indd 249 29/04/2013 11:51
and temporal intelligence can now work together to express what
we already knew but could not communicateas well as generate
new messages and experiences whose meanings we have yet to
Although we may still need to gure out how to fully use this
new semiotic meta-language, the importance of its emergence is
hard to overestimate. In short, the emergence of software-enabled
motion graphics is as important historically as the invention of
printing, photography, or the Internet.
We will begin by going back to the 1980s. During the heyday
of post-modern debates, at least one critic in America noticed the
connection between post-modern pastiche and computerization. In
his book After the Great Divide (1986), Andreas Huyssen wrote,
All modern and avantgardist techniques, forms and images are now
stored for instant recall in the computerized memory banks of our
culture. But the same memory also stores all of pre-modernist art as
well as the genres, codes, and image worlds of popular cultures and
modern mass culture.
His analysis is accurateexcept that these
computerized memory banks did not really become commonplace
for another fteen years. Only when the Web absorbed enough of
the media archives did it become a universal cultural memory bank
accessible to all cultural producers. But even for the professionals,
the ability to easily integrate multiple media sources within the same
projectmultiple layers of video, scanned still images, animation,
graphics, and typographyonly came towards the end of the 1990s.
In 1985, when Huyssens book was in preparation for publi-
cation, I was working for one of the few computer animation
companies in the world. The company was located in NYC and it
was appropriately called Digital Efects.
Each computer animator
had his/her own interactive graphics terminal that could show 3D
models but only in wireframe and monochrome; to see them fully
rendered in color, we had to take turns as the company had only
one color raster display which we all shared. The data was stored
on bulky magnetic tapes about a foot in diameter; to nd the data
Andreas Huyssen, Mapping the Postmodern, in After the Great Divide
(Bloomington and Indianapolis: Indiana University Press, 1986), p. 196.
Wayne Carlson, A Critical History of Computer Graphics and
Animations. Section 2: The Emergence of Computer Graphics Technology,
9781623568177_txt_print.indd 250 29/04/2013 11:51
from an old job was a cumbersome process that involved locating
the right tape in the tape library, putting it on a tape drive and then
searching for the right part of the tape. We did not have a color
scanner, so getting all modern and avantgardist techniques, forms
and images into the computer was far from trivial. And even if we
had had one, there was no way to store, recall, and modify these
images. The machine that could do thatQuantel Paintboxcost
over $160,000, which we could not aford. And when in 1986
Quantel introduced Harry, the rst commercial non-linear editing
system that allowed for digital compositing of multiple layers of
video and special efects, its cost similarly made it prohibitive for
everybody except network television stations and a few production
houses. Harrys capacities were quite limited, because it could
record only eighty seconds of broadcast quality video. In the realm
of still images, things were not much better: for instance, the digital
still store Picturebox released by Quantel in 1990 could hold only
500 broadcast quality images and its cost was similarly very high.
In short, in the middle of the 1980s neither we nor other
production companies had anything approaching the comput-
erized memory banks imagined by Huyssen. And of course, the
same was true for the visual artists that were associated with
post-modernism and the ideas of pastiche, collage, and appro-
priation. In 1986 the BBC produced a documentary Painting
with Light for which half a dozen well-known painters including
Richard Hamilton and David Hockney were invited to work with
the Quantel Paintbox. The resulting images were not so diferent
from the normal paintings that these artists were producing
without a computer. And while some artists were making reference
to modern and avantgardist techniques, forms and images,
these references were painted rather than being directly loaded
from computerized memory banks. Only about ten years later,
when relatively inexpensive graphics workstations and personal
computers running image editing, animation, compositing, and
illustration software became commonplace and afordable for
freelance graphic designers, illustrators, and small post-production
and animation studios did the situation described by Huyssen start
to become a reality.
The results were dramatic. Within the space of less than ve
years, modern visual culture was fundamentally transformed.
Visuals which previously were specic to diferent medialive
9781623568177_txt_print.indd 251 29/04/2013 11:51
action cinematography, graphics, still photography, animation, 3D
computer animation, and typographystarted to be combined in
numerous ways. By the end of the 1990s, the pure moving image
media became an exception and hybrid media became the norm.
However, in contrast to other computer revolutions such as the
rise of the World Wide Web around the same time, this revolution
was not acknowledged by popular media or cultural critics.
What received attention were the developments that afected
narrative lmmakingthe use of computer-produced special efects
in Hollywood feature lms or the inexpensive digital video and
editing tools outside of it. But another process which happened
on a larger scalethe transformation of the visual language used
by all forms of moving images outside of narrative lmshas not
been critically analyzed. In fact, while the results of these transfor-
mations have become fully visible by 1999, at the time of writing I
am not aware of a single theoretical article discussing them (at least
in English).
One of the reasons is that in this revolution no new media per
se were created. Just as they did ten years ago, the designers were
making still images and moving images. But the aesthetics of these
images was now very diferent. In fact, it was so new that, in retro-
spect, the post-modern imagery of just ten years ago that at the
time looked strikingly diferent now appears as a barely noticeable
blip on the radar of cultural history.
My choice of the starting and ending dates (19939) to characterize
the development of a new hybrid visual language of moving images
is not accidental. Of course, I could have picked diferent datesfor
instancestarting a few years earlierbut since After Efects software
which will play the key role in my account was released in 1993, I
decided to pick this year as my starting date. And while my ending
date also could have been diferent, I believe that by 1999 the broad
changes in the aesthetics of moving images became visible. If you
want to quickly see this for yourself, simply compare demo reels from
the same visual efects companies made in the early 1990s and the late
1990s (a number of them are available onlinelook, for instance, at
the demo reels of Pacic Data Images, or the demo reels of the Flame
system, available for every year starting in 1995.
) In work from the
http://accad.osu.edu/~waynec/history/lesson6.html; http://area.autodesk.com/ flame20
9781623568177_txt_print.indd 252 29/04/2013 11:51
beginning of the decade, computer imagery in most cases appears
by itselfthat is, we see whole commercials and promotional videos
done in 3D computer animation, and the novelty of this new media is
foregrounded. By the end of the 1990s, computer animation becomes
just one element integrated in the media mix that also includes live
action, typography, and design.
Although these transformations happened only recently, the
ubiquity of the new hybrid visual language today is such that it takes
an efort to recall how diferent things looked before. Similarly, the
changes in production processes and equipment that made this
language possible also quickly fade from public and professional
memories. As a way to quickly evoke these changes as seen from the
professional perspective, I am going to quote from a 2004 interview
with Mindi Lipschultz, who has worked as an editor, producer, and
director in Los Angeles since 1979:
If you wanted to be more creative [in the 1980s], you couldnt
just add more software to your system. You had to spend
hundreds of thousands of dollars and buy a paintbox. If you
wanted to do something graphican open to a TV show with a
lot of layersyou had to go to an editing house and spend over
a thousand dollars an hour to do the exact same thing you do
now by buying an inexpensive computer and several software
programs. Now with Adobe After Efects and Photoshop, you
can do everything in one sweep. You can edit, design, animate.
You can do 3D or 2D all on your desktop computer at home or
in a small ofce.
In 1989, the former Soviet satellites of Central and Eastern
Europe peacefully liberated themselves from the Soviet Union.
In the case of Czechoslovakia, this event came to be referred as
the Velvet Revolutionto contrast it to typical revolutions in
modern history that were always accompanied by bloodshed. To
emphasize the gradual, almost invisible pace of the transformations
which occurred in moving image aesthetics between approximately
1993 and 1999, I am going to appropriate the term the Velvet
Revolution to refer to these transformations. (Although it may
Mindi Lipschultz, interviewed by The Compulsive Creative, May 2004, http://
9781623568177_txt_print.indd 253 29/04/2013 11:51
seem presumptuous to compare political and aesthetics transfor-
mations simply because they share the same non-violent quality, it
is possible to show that the two revolutions are actually related.)
Finally, before proceeding I should also explain my use of
examples. The visual language I am analyzing is all around us
today (this may explain why academics have remained blind to
it). After globalization, this language is spoken by communi-
cation professionals in dozens of countries around the world. You
can see for yourself all the examples of various aesthetics I will
be mentioning below by simply watching television and paying
attention to graphics, going to a club to see a VJ performance,
visiting the websites of motion graphics designers and visual
efects companies, or opening any book on contemporary design.
Nevertheless, below I have included titles of particular projects so
the reader can see exactly what I am referring to. (I have chosen
works by well-known design studies and artists so you can easily
nd all of them on the web.) But since my goal is to describe
the new cultural language that by now has become practically
universal, I want to emphasize that each of these examples can be
substituted with numerous others.
The aesthetics of hybridity
In the second half of the 1990s, one of the key identifying features
of motion graphics that clearly separated it from other forms of
moving image existing until that time, was the central role played
by dynamic typography. The term motion graphics has been used
at least since 1960, when a pioneer of computer lmmaking John
Whitney named his new company Motion Graphics. However, until
the Velvet Revolution only a handful of people and companies have
systematically explored the art of animated typography: Norman
McLaren, Saul Bass, Pablo Ferro, R/Greenberg, and a few others.

But by the middle of the 1990s moving image sequences or short
For a rare discussion of motion graphics prehistory as well as an equally rare
attempt to analyze the eld by using a set of concepts rather than only presenting
the portfolio of individual designers, see Jef Bellantoni and Matt Woolman, Type
in Motion, 2
edition (Thames & Hudson, 2004).
9781623568177_txt_print.indd 254 29/04/2013 11:51
lms dominated by moving animated type and abstract graphical
elements rather than by live action, started to be produced in large
numbers. What was the material cause for motion graphics taking
of? After Efects and other related software running on PCs or
relatively inexpensive graphics workstations became afordable
to smaller design, visual efects, and post-production houses, and
soon, to individual designers. Almost overnight, the term motion
graphics became well known. (As the Wikipedia article about this
term points out, the term Motion Graphics was popularized by
Trish and Chris Meyers book about the use of Adobe After Efects
titled Creating Motion Graphics.
) The ve-hundred-year-old
Gutenberg universe came into motion.
Along with typography, the whole language of twentieth-century
graphic design was imported into moving image design. While
this development did not receive a popular new name of its own,
it is obviously at least as important. (Although the term design
cinema has been used, it never achieved anything comparable to
the popularity of motion graphics.) So while motion graphics
were for years limited to lm titles and therefore only used typog-
raphy, today the term motion graphics is often used to refer to
moving image sequences that combine moving type and design
elements. But we should recall that while in the twentieth century
typography was often used in combination with other design
elements, for 500 years it commanded its own word. Therefore I
think it is important to consider the two kinds of import opera-
tions that took place during the Velvet Revolutiontypography
and twentieth century graphic designas two distinct historical
While motion graphics denitely exemplify the changes that
took place during the Velvet Revolution, these changes are broader.
Simply put, the result of the Velvet Revolution is a new hybrid
visual language of moving images in general. This language is not
conned to particular media forms. And while today it manifests
itself most clearly in non-narrative forms, it is also often present in
narrative and gurative sequences and lms.
Here are a few examples. A music video may use live action
while also employing typography and a variety of transitions done
9781623568177_txt_print.indd 255 29/04/2013 11:51
Sony Reader. D-FUSE, 2009. Selected frames from a motion graphics
9781623568177_txt_print.indd 256 29/04/2013 11:51
with computer graphics (video for Go by Common, directed
by Convert/MK12/Kanye West, 2005). Another music video may
embed the singer within an animated painterly space (video for
Sheryl Crows Good Is Good, directed by Psyop, 2005). A
commercial may superimpose charts, data displays, and data
visualizations on top of live action (TV ad for Thomson Reuters
by MK12, 2012). The title sequence may contrast 2D at gures
and deep 3D perspectival space (titles for Mad Men by Imaginary
Forces, 2007). A short lm may mix typography, stylized 3D
graphics, animated 2D design elements, and live action (Itsu for
Plaid, directed by the Pleix collective, 2002). (Sometimes, as I have
already mentioned, the term design cinema is used to difer-
entiate such short independent lms organized around design,
typography, and computer animation rather than live action from
similar motion graphics works produced for commercial clients.)
In some cases, the juxtaposition of diferent media is clearly
visible (video for Dont Panic by Coldplay, 2001; titles for the
television show The Inside, 2005; commercial Nike Dynamic
Feet, 2005, all by Imaginary Forces). In other cases, a sequence
may move between diferent media so quickly that the shifts are
barely noticeable (GMC Denali Holes commercial by Imaginary
Forces, 2005). Yet in other cases, a commercial or movie title may
feature continuous action shot on video or lm, with the image
periodically changing from a more natural to a highly stylized look.
Such media hybridity does not necessarily manifest itself in
a collage-like aesthetics that foregrounds the juxtaposition of
diferent media and diferent media techniques. As a very diferent
example of what media hybridity can result in, consider a more
subtle aesthetics well captured by the name of the software that
to a large extent made the hybrid visual language possible: After
Efects. This name anticipated the changes in visual efects that
took place a number of years later. In the 1990s computers were
used to create highly spectacular special efects or invisible
but towards the end of that decade we see something
Invisible efect is the standard industry term. For instance, the lm Contact,
directed by Robert Zemeckis, was nominated for the 1997 VFX HQ Awards in
the following categories: Best Visual Efects, Best Sequence (The Ride), Best Shot
(Powers of Ten), Best Invisible Efects (Dish Restoration), and Best Compositing.
9781623568177_txt_print.indd 257 29/04/2013 11:51
else emerging: a new visual aesthetics that goes beyond efects.
In this aesthetics, the whole projectwhether a music video, a TV
commercial, a short lm, or a large segment of a feature lmhas
a special look in which the enhancement of live-action material is
not completely invisible but at the same time does not call attention
to itself the way special efects tended to do in the 1990s (examples:
Reebok I-Pump Basketball Black commercial and The Legend of
Zorro main title, both by Imaginary Forces, 2005; Fage Plain
commercial by Psyop, 2011).
Although the particular aesthetic solutions vary from one video
to the next and from one designer to another, they share the same
logic: the simultaneous appearance of multiple media within the
same frame. Whether these media are openly juxtaposed or almost
seamlessly blended together is less important than the fact of this
copresence itself. (Again, note that each of the examples above can
be substituted with numerous others.)
Hybrid visual language is also now common to a large proportion
of short experimental and independent (i.e., not commis-
sioned by commercial clients) videos being produced for media
festivals, the web, mobile media devices, and other distribution
Many visuals created by VJs and live cinema artists
are also hybrid, combining video, layers of 2D imagery, animation,
and abstract imagery generated in real time.
And as the anima-
tions of artists Jeremy Blake, Ann Lislegaard, and Takeshi Murata
that I will discuss below demonstrate, at least some of the works
created explicitly for art-world distribution similarly choose to use
the same language of hybridity.
In December 2005, I attended the Impact media festival in Utrecht and asked the
festival director what percentage of the submissions they received that year featured
hybrid visual language as opposed to straight video or lm. His estimate was
about 50 percent. In January 2006, I was part of the review team that judged the
projects of students graduating from SCI-ARC, a well-known research-oriented
architecture school in Los Angeles. According to my informal estimate, approxi-
mately half the projects featured complex curved geometry made possible by Maya,
a modeling software now commonly used by architects. Given that both After
Efects and Mayas predecessor, Alias, were introduced in the same year1993I
nd this quantitative similarity in the percentage of projects that use new languages
made possible by this software quite telling.
For examples, consult Paul Spinrad, ed., The VJ Book: Inspirations and Practical
Advice for Live Visuals Performance (Feral House, 2005).
9781623568177_txt_print.indd 258 29/04/2013 11:51
Today, narrative features rarely mix diferent graphical styles
within the same frame. However, gradually growing number of
lms do feature highly stylized aesthetics that would have previ-
ously been identied with illustration rather than lmmaking.
The examples are the Wachowskis Matrix series (19992003),
Immortal by Enki Bilal (2004), Robert Rodriguez Sin City (2005)
and The Spirit (2008), Zack Snyders 300 (2007) and Watchmen
(2009), James Camerons Avatar (2009), Tim Burtons Alice In
Wonderland (2010) and Martin Scorseses Hugo (2011). These
feature lms are examples of now fully established practice to
shoot a large portion of a feature lm using a digital backlot
(i.e., a green screen).
Consequently, most or all shots in such lms
are created by composing the footage of actors with computer-
generated sets and other visuals.
These lms do not juxtapose their diferent media as dramati-
cally as motion graphics. Nor do they strive for the seamless
integration of CGI (computer-generated imagery) visuals and live
action that characterized the earlier special-efects features of the
1990s, such as Terminator 2 (1991) and Titanic (1997) by James
Cameron. Instead, they explore the space in between juxtaposition
and complete integration.
Matrix, Sin City, 300, Alice In Wonderland, Hugo, and other
lms shot on a digital backlot combine multiple media to create
a new stylized aesthetics that cannot be reduced to the already
familiar look of live-action cinematography or 3D computer
animation. Such lms display exactly the same logic as short
motion graphics works, which at rst sight might appear to be
very diferent. This logic is also the same as that which we observe
in the creation of new hybrids in biology. That is, the result of the
hybridization process is not simply a mechanical sum of the previ-
ously existing parts but a new speciesa new kind of visual
aesthetics that did not exist previously.
In TV commercials produced in the 2000s, this highly stylized
visual aesthetics became one of the key looks of the decade. Many
layers of live footage, 3D and 2D animated elements, particle
efects, and other media elements are blended to create a seamless
whole. This result has the crucial codes of realism (perspective
9781623568177_txt_print.indd 259 29/04/2013 11:51
foreshortening, atmospheric perspective, correct combination of
lights and shadows), but at the same time enhances visible reality.
(I cannot call this aesthetics hyperreal since the hybrid images
assembled from many layers and types of media look quite diferent
from the works of the hyperrealist artists such as Denis Peterson that
visually look like standard color photographs.) Strong gray-scale
contrast, high color saturation, tiny waves of particles emulating
from moving objects, extreme close-ups of textured surfaces (water
drops, food products, human skin, nishes of consumer electronics
devices, etc.), the contrasts between the natural uneven textured
surfaces and smooth 3D renderings and 2D gradients, the rapidly
changing composition and camera position and direction, and
other devices heighten our perception. (For examples of all these
strategies, you can, for example, look at the commercials made by
We can say that these commercials create a map which is bigger
than the territory being mapped, because they show more details
and texture spatially, and at the same time compress time, moving
through information more rapidly. We can also make a comparison
with the Earth observation satellites which circle the planet, capturing
its whole surface in detail impossible for any human observer to
seejust as a human being cannot simultaneously see the extreme
close-up of the surfaces and details of the movements of objects
presented in the ctional space of a commercial.
In summary, the result of the shift to a software production
environment in moving image creation is a new visual language.
Just as with the individual software techniques that make it
possible, this language as a whole inherits the traits of previous
image medialmmaking, cel and puppet animation, computer
animation, photography, painting, graphic design and typography.
However, it is not reducible to any of these media. Rather, it is a
true hybridthe ofspring of twentieth-century image mediums
that shares common traits with all of them but has its own distinct
Let us now look at two short lms in detail to see how aesthetics
in hybridity works across a whole lm. Blakes Sodium Fox (2005)
and Muratas Untitled (Pink Dot) (2007) ofer excellent examples
http://www.psyop.tv/projects/live-action/ (October 31, 2012)
9781623568177_txt_print.indd 260 29/04/2013 11:51
of the new hybrid visual language that currently dominates moving-
image culture. Among many well-known artists working with
digital moving images, Blake was the earliest and most successful
in developing his own style of hybrid media. His video Sodium
Fox is a sophisticated blend of drawings, paintings, 2D animation,
photography, and efects available in software. Using a strategy
commonly employed by artists in relation to commercial media
in the twentieth century, Blake slows down the fast-paced rhythm
of motion graphics as they are usually practiced today. However,
despite the seemingly slow pace of his lm, it is as informationally
dense as the most frantically changing motion graphics used in clubs,
music videos, television station IDs, and so on. Sodium Fox creates
this density by exploring in an original way the basic feature of the
software-based production environment in general and programs
such as After Efects in particularnamely, the construction of
an image from potentially numerous layers. Of course, traditional
cel animation as practiced in the twentieth century also involved
building up an image from a number of superimposed transparent
cels, with each one containing some of the elements that together
make up the whole image. For instance, one cel could contain a
face, another lips, a third hair, yet another a car, and so on.
With computer software, however, designers can precisely
control the transparency of each layer; they can also add diferent
visual efects, such as blur, between layers. As a result, rather than
creating a visual narrative based on the motion of a few visual
elements through space (as was common in twentieth-century
animation, both commercial and experimental), designers now
have many new ways to create animation. Exploring these possibil-
ities, Blake crafts his own visual language in which visual elements
positioned on diferent layers are continuously and gradually
written over each other. If we connect this new language to
twentieth-century cinema rather than to cel animation, we can say
that rather than fading in a new frame as a whole, Blake continu-
ously fades in separate parts of an image. The result is an aesthetics
that balances visual continuity with a constant rhythm of visual
rewriting, erasing, and gradual superimposition.
Like Sodium Fox, Muratas Untitled (Pink Dot) also develops
its own language within the general paradigm of media hybridity.
Murata creates a pulsating and breathing image that has a distinctly
biological feel to it. In the last decade, many designers and artists
9781623568177_txt_print.indd 261 29/04/2013 11:51
Sodium Fox. Jeremy Blake, 2005. Selected frames from a 14-minute
digital video.
9781623568177_txt_print.indd 262 29/04/2013 11:51
9781623568177_txt_print.indd 263 29/04/2013 11:51
Untitled (Pink Dot). Takeshi Murata, 2007. Selected frames from a ve-
minute digital video.
9781623568177_txt_print.indd 264 29/04/2013 11:51
9781623568177_txt_print.indd 265 29/04/2013 11:51
have used biologically inspired algorithms and techniques to create
animal-like movements in their generative animations and interac-
tives. However, in the case of Untitled (Pink Dot), the image as a
whole seems to come to life.
To create this pulsating, breathing-like rhythm, Murata trans-
forms live-action footage (scenes from 1982 movie Rambo: First
Blood) into a ow of abstract color patches (sometimes they look
like oversize pixels, and at other times they may be taken for
artifacts of heavy image compression). But this transformation
never settles into a nal state. Instead, Murata constantly adjusts
its degree. (In terms of the interfaces of media software, this would
correspond to animating a setting of a lter or an efect.) One
moment we see almost unprocessed live imagery; the next moment
it becomes a completely abstract pattern; in the following moment,
parts of the live action image again become visible, and so on.
In Untitled (Pink Dot) the general condition of media hybridity
is realized as a permanent metamorphosis. True, we still see some
echoes of movement through space, which was the core method
of pre-digital animation. (Here this is the movement of the gures
in the shots from Rambo.) But now the real change that matters
is the one between diferent media aesthetics: between the texture
of a lm and the pulsating abstract patterns of owing patches of
color, between the original liveness of human gures in action as
captured on lm and the highly exaggerated articial liveness they
generate when processed by a machine.
Visually, Untitled (Pink Dot) and Sodium Fox do not have much
in common. However, both lms share the same strategy: creating
a visual narrative through continuous transformations of image
layers, as opposed to discrete movements of graphical marks or
characters, common to both the classic commercial animation of
Disney and the experimental classics of Norman McLaren, Oskar
Fischinger, and others. Although we can assume that neither Blake
nor Murata has aimed to achieve this consciously, in diferent ways
each artist stages for us the key technical and conceptual change
that denes the new era of media hybridity. Media software allows
the designer to combine any number of visual elements regardless
of their original media and to control each element in the process.
This basic ability can be explored in numerous visual aesthetics.
The lms of Blake and Murata, with their diferent temporal
rhythms and diferent logics of media combination, exemplify
9781623568177_txt_print.indd 266 29/04/2013 11:51
this diversity. Blake layers various still graphics, text, animation,
and efects, dissolving elements in and out. Murata processes live
footage to create a constant image ow in which the two layers
live footage and its processed resultseem to constantly push each
other out.
Deep remixability
I believe that media hybridity constitutes a new fundamental stage
in the history of media. It manifests itself in diferent areas of software
culture and not only in moving imagesalthough the latter does ofer
a particularly striking example of this new cultural logic at work.
Here the media authoring software environment became a kind of
Petri dish where the techniques and tools of computer animation, live
cinematography, graphic design, 2D animation, typography, painting
and drawing can interact, generating new hybrids. And as the
examples above demonstrate, the results of this process of hybridity
are a new aesthetics and new media species that cannot be reduced
to the sum of the media that went into their creation.
Can we understand the new hybrid language of moving image as
a type of remix? From its beginnings in music culture in the 1980s,
during the 1990s remix has gradually emerged as the dominant
aesthetics of the era of globalization, afecting and re-shaping
everything from music and cinema to food and fashion. (If Fredric
Jameson once referred to post-modernism as the cultural logic of
late capitalism, we can perhaps call remix the cultural logic of
networked global capitalism.) A number of authors have already
traced remix efects in many cultural areas, ranging from childrens
use of media in Japan (Mimi Ito) to web culture (Eduardo Navas).
The representative books include Rhythm Science (D. J. Spooky,
2004), Remix: Making Art and Commerce Thrive in the Hybrid
Economy (Lawrence Lessig, 2008), Mashed Up: Music, Technology,
The Rise of Congurable Culture (Aram Sinnreich, 2010), Mashup
Cultures (edited by Stefan Sonvilla-Weiss, 2010) and Remix Theory:
The Aesthetics of Sampling (Eduardo Navas, 2012).
Mizuko Ito, Mobilizing the Imagination in Everyday Play: The Case of Japanese
Media Mixes, in International Handbook of Children, Media, and Culture,
9781623568177_txt_print.indd 267 29/04/2013 11:51
I believe that the combinatory mechanisms responsible for the
evolution of the computer metamedium in general, and the
new hybrid visual aesthetics emerging in the 1990s can indeed be
considered as a type of remixif we make one crucial distinction.
Typical remix combines content within the same media or content
from diferent media. For instance, a music remix may combine
music elements from any number of artists; anime music videos
may combine parts of anime lms and music taken from a music
video. Professionally produced motion graphics and other moving-
image projects also routinely mix together content in the same
media and/or from diferent media. For example, in the beginning
of the Go music video (Convert/MK12/Kanye West, 2005), the
video rapidly switches between live-action footage of a room and a
3D model of the same room. Later, the live-action shots also incor-
porate a computer-generated plant and a still photographic image
of mountain landscape. Shots of a female dancer are combined
with elaborate animated typography. The human characters are
transformed into abstract animated patterns. And so on.
Such remixes of content from diferent media are denitely
common today in moving-image culture. In fact, I began discussing
the new visual language by pointing out that in the case of short
forms they now constitute the rule rather than the exception. But
this type of remix is only one aspect of the hybrid revolution.
For me its essence lies in something else. Let us call it deep remix-
ability. Today designers remix not only content from diferent
media but also their fundamental techniques, working methods,
and ways of representation and expression. United within the
common software environment the languages of cinematography,
animation, computer animation, special efects, graphic design,
typography, drawing, and painting have come to form a new
metalanguage. A work produced in this new metalanguage can
Sonia Livingstone and Kirsten Drotner, (eds) (Sage Publications, 2008); Paul D.
Miller, Rhythm Science (MIT Press, 2004); Lawrence Lessig. Remix: Making Art
and Commerce Thrive in the Hybrid Economy (Penguin Press HC, 2008); Aram
Sinnreich, Mashed Up: Music, Technology, and the Rise of Congurable Culture
(University of Massachusetts Press, 2010); Stefan Sonvilla-Weiss, Mashup Cultures
(Springer, 2010); Eduardo Navas, Remix Theory: The Aesthetics of Sampling
(Springer Vienna Architecture, 2012).
9781623568177_txt_print.indd 268 29/04/2013 11:51
use all the techniques, or any subset of these techniques, that were
previously unique to these diferent media.
We may think of this new metalanguage of moving images as
a large library of all previously known techniques for creating
and modifying moving images. A designer of moving images
selects techniques from this library and combines them in a single
sequence or a single frame. But this clear picture is deceptive. How
exactly does s/he combine these techniques? When you remix
content, it is easy to imagine diferent texts, audio samples, visual
elements, or data streams positioned side by side. Imagine a typical
twentieth-century collage except that it now moves and changes
over time. But how do you remix the techniques themselves?
In the cases of hybrid media interfaces that we have already
analyzed (such as Acrobats interface), remix of techniques
means simple combination. Diferent techniques literally appear
next to each in the applications UI. Thus, in Acrobat, a forward
and backward button, a zoom button, a nd tool, and others are
positioned one after another on a toolbar above the open document.
Other techniques appear as tools listed in vertical pull-down menus:
spell, search, email, print, and so on. We nd the same principles in
the interfaces of all media authoring and access applications. The
techniques borrowed from various media and the new born-digital
techniques are presented side-by-side using tool-bars, pull-down
menus, toolboxes and other UI design conventions.
Such an addition of techniques that exist in a single space side
by side without any deep interactions are also indirectly present
in remixes of content well familiar to us, be it fashion designs,
architecture, collages, or motion graphics. Consider a hypothetical
example of a visual design that combines drawn elements, photos,
and 3D computer graphics forms. Each of these visual elements
is the result of the use of particular media techniques of drawing,
photography, and computer graphics. Thus, while we may refer to
such cultural objects as remixes of content, we are also justied in
thinking about them as remixes of techniques. This applies equally
well to pre-digital design, when a designer would use separate
physical tools or machines, and to contemporary software-driven
design, where s/he has access to all these tools in a few compatible
software applications.
As long as the pieces of content, interface buttons, or techniques
are simply added rather than integrated together, we do not need a
9781623568177_txt_print.indd 269 29/04/2013 11:51
Music video for Go! By Common. Kanye West, MK12 and Convert,
2005. Selected frames from a four-minute video. A number below each
frame indicates time code (in seconds) for this frame.
9781623568177_txt_print.indd 270 29/04/2013 11:51
9781623568177_txt_print.indd 271 29/04/2013 11:51
special term such as deep remix. This, for me, is still remix the
way this term is commonly used. But in the case of moving image
aesthetics we also encounter something else. Rather than a simple
addition, we also nd interactions between previously separate
techniques of cel animation, cinematography, 3D animation,
design, and so oninteractions which were unthinkable before.
(The same argument can be made in relation to other types of
cultural objects and experiences created with media authoring
software such as visual designs and music.)
I believe that this is something that neither pioneers of computer
media of the 1960s1970s nor the designers of the rst media
authoring applications that started to appear in the 1980s intended.
However, once all these media techniques met within the same
software environmentand this was gradually accomplished
throughout the 1990sthey started interacting in ways that could
never have been predicted or even imagined previously.
For instance, while particular media techniques continue to be
used in relation to their original media, they can also be applied
to other media. Photoshop lters, which we previously analyzed,
illustrate well this cross-over efect: techniques that were origi-
nally part of a particular media type can now be applied to other
media types. For instance, neon glow, stained glass, lens are, etc.
can be applied to photographs and sketches. (More precisely, they
can be applied to whatever is currently loaded in the graphics
memory and appears in the image windowa set of pixels
that carry the results of all previously applied lters and other
Here are typical examples of the crossover strategy as it is used
in moving image design. Type is choreographed to move in 3D
space; motion blur is applied to 3D computer graphics; algorith-
mically generated elds of particles are blended with live-action
footage to give them an enhanced look; a virtual camera is made
to move around a virtual space lled with 2D drawings. In each
of these examples, the technique that was originally associated
with a particular mediumcinema, cel animation, photorealistic
computer graphics, typography, graphic designis now applied to
a diferent media type. Today a typical short lm or a sequence may
combine many of such pairings within the same frame. The result
is a hybrid, intricate, complex, and rich media languageor rather,
numerous languages that share the logic of deep remixability.
9781623568177_txt_print.indd 272 29/04/2013 11:51
In fact, such interactions among virtualized media techniques
dene the aesthetics of contemporary moving image culture. This is
why I have decided to introduce a special termdeep remixability.
I wanted to diferentiate more complex forms of interactions
between techniques (such as cross over) from the simple remix (i.e.
addition) of media content and media techniques with which we
are all familiar, be it music remixes, anime video remixes, 1980s
postmodern art and architecture, and so on.
For concrete examples of the crossover efect which exemplies
deep remixability, we can return to the same Go video and look
at it again, but now from a new perspective. Previously I have
pointed out the ways in which this videotypical of the short
format moving images works todaycombines visual elements of
diferent media types: live action video, still photographs, proce-
durally generated elements, typography, etc. However, exactly the
same shots also contain rich examples of the interactions between
techniques, which are only possible in a software-driven design
As the video begins, a structure made up from perpendicular
monochrome blocks and panels simultaneously grows rapidly
in space and rotates to settle into a position which allows us
to recognize it as a room (00:0700:11). As this move is being
completed, the room is transformed from an abstract geometric
structure into a photo-realistically rendered one: furniture pops in,
wood texture rolls over the oor plane, and a photograph of a
mountain view lls a window. Although such diferent styles of CG
rendering have been available in animation software since the 1980s,
a particular way in which this video opens with a visually striking
abstract monochrome 3D structure is a clear example of deep remix-
ability. When in the middle of the 1990s graphic designers started
to use computer animation software, they brought their training,
techniques, and sensibilities to computer animation that until that
time was used in the service of photorealism. The strong diagonal
compositions, the deliberately at rendering, and the choice of colors
in the opening of the Go video subordinates CG photorealistic
techniques to a visual discipline specic to modern graphic design.
The animated 3D structure references the Suprematism of Malevich
and Lissitzky, which played a key role in shaping the grammar
of modern designand which, in our example, has become a
conceptual lter that has transformed the CG eld.
9781623568177_txt_print.indd 273 29/04/2013 11:51
After a momentary stop to let us take in the room, which is now
largely completed, a camera suddenly rotates 90 degrees (00:15
00:17). This physically impossible camera move is another example
of deep remixability. While animation software implements the
standard grammar of twentieth-century cinematographya pan,
a zoom, a dolly, etc.the software, of course, does not have the
limitations of a physical world. Consequently a camera can move
in an arbitrary direction, follow any imaginable curve and do this
at any speed. Such impossible camera moves become standard tools
of contemporary media design and twenty-rst-century cinematog-
raphy, appearing with increased frequency in feature lms since the
middle of the 2000s. Just as Photoshop lters which can be applied
to any visual composition, virtual camera moves can also be super-
imposed, so to speak, on any visual scene regardless of whether it
was constructed in 3D, procedurally generated, captured on video,
photographed, or drawnor, as in the example of the room from
Go video, is a combination of these diferent media.
Playing the video forward (00:1500:22), we notice yet another
previously impossible interaction between media techniques.
The interaction in question is a lens reection, which is slowly
moving across the whole scene. Originally an artifact of a camera
technology, lens reection was turned into a lteri.e., a technique
which can now be drawn over any image constructed with all
other techniques available to a designer. (This important type of
software technique, one which originated as artifacts of physical or
electronic media technologies, will be discussed in more details in
the concluding section of this chapter.) If you wanted more proof
that we are dealing here with a visual technique, note that this
lens reection is moving while the camera remains perfectly still
(00:1700:22)a logical impossibility, which is sacriced in favor
of a more dynamic visual experience.
I referred to the new language of moving imagery as a meta-
language. Since our discussion has so far relied on the term
computer metamedium, I should explain the connection between
these two terms.
The acceleration of the speed of social, technological and
cultural change in the second part of the twentieth century has
led to the frequent use of meta-, hyper-, and super- in cultural
theory and criticism. From Superstudio (a conceptual architec-
tural group active in the 1960s), Ted Nelsons Hypermedia and
9781623568177_txt_print.indd 274 29/04/2013 11:51
Alan Kays metamedium to the more recent Supermodernism and
these terms may be read as attempts to capture
the feeling that we have passed a point of singularity and are
now moving at warp speed. Like the cosmonauts of the 1960s
observing the Earth from the orbits of their spaceships and seeing
it for the rst time as a single object, we are looking down at
human history from a new higher orbit. This connotation seems
to t Alan Kays conceptual and practical redenition of a digital
computer as a metamedium that contains most existing medium
technologies and techniques and also allows invention of many
new ones.
While the term metalanguage has precise meanings in logic,
linguistics, and computing, here I am using it in a sense similar to
Alan Kays use of meta in computer metamedium. Normally
a metalanguage refers to a separate formal system for describing
mediums or cultural languagesthe way grammar describes how
a particular natural language works. But this not how Kay uses
meta in metamedium. As he uses it, it stands for gathering/
including/collectingin short, bringing previously separate things
Let us imagine this computer metamedium as a large and
continuously expanding set of resources. It includes all media
creation/manipulation techniques, interaction techniques and data
formats available to programmers and designers in the current
historical moment. Everything from sort and search algorithms and
pull-down menus to hair and water rendering techniques, video
games AI, and multi-touch interface methodsit is all there.
If we look at how these resources are used in diferent cultural
areas to create particular kinds of content and experiences, we
will see that each of them only uses a subset of these resources.
For example, the graphical interfaces of todays popular computer
operating systems (Windows, Linux, Mac OS) use static icons. In
contrast, in some consumer electronics interfaces (such as certain
mobile phones) all icons are animated loops.
Moreover, the use of a subset of all existing elements is not
random but follows particular conventions. Some elements always
go together. In other cases, the use of one element means that we
9781623568177_txt_print.indd 275 29/04/2013 11:51
are unlikely to nd some other element. In other words, diferent
forms of digital media use diferent subsets from a complete set of
techniques contained in a computer metamediumand this use
follows distinct patterns.
If you notice the parallels with what cultural critics usually call
an artistic language, a style, or a genre, you are right. Any
single work of literature or works of a particular author or literary
movement uses only some of the existing literary techniques, and this
use follows some patterns. The same goes for cinema, music and all
other recognized cultural forms. This allows us to talk about a style
of a particular novel or lm, a style of an author as a whole, or a
style of a whole artistic school. (Film scholars David Bordwell and
Kristin Thompson call this a stylistic system which they dene as
a patterned and signicant use of techniques. For cinema, they
divide these techniques into four categories: mise-en-scne, cinema-
tography, editing, and sound.
) When a whole cultural eld can
be divided into a small number of distinct groups of works with
each group sharing some patterns, we usually talk about genres.
For instance, theoreticians of Ancient Greek theatre distinguished
between comedies and tragedies and prescribed the rules each genre
should follow, while today companies use automatic software to
classify blogs into diferent genres.
If by medium we mean a set of standard technological resources,
be it a physical stage or a lm camera, lights and lm stock,
we can see that each medium usually supports multiple artistic
languages/styles/genres. For example, a medium of twentieth-
century lmmaking supported Russian Montage of the 1920s,
Italian Neorealism of the 1940s, French New Wave of the 1960s,
Hong Kong fantasy Kung Fu lms of the 1980s, Chinese fth-
generation lms of the 1980s1990s, etc.
Similarly, a computer metamedium can support multiple cultural
or artistic metalanguages. In other words, in the theoretical
scheme I am proposing, there is only one metamediumbut many
So what is a metalanguage? If we dene an artistic language as
a patterned use of a selected number of a subset of the techniques
David Bordwell and Kristin Thompson, Film Art: an Introduction, 5th edition
(The McGraw-Hill Companies, 1997), p. 355.
9781623568177_txt_print.indd 276 29/04/2013 11:51
available in a given medium,
a metalanguage is a patterned
use of a subset of all the techniques available in a computer
metamedium. But not just any subset. It only makes sense to talk
about a metalanguage (as opposed to a language) if the techniques
it uses come from previously distinct cultural languages. As an
example, consider a metalanguage of popular commercial virtual
globes (Google Earth, Microsoft Bing Maps). These applica-
tions systematically combine diferent types of media formats and
media navigation techniques that previously were separate. These
combinations follow common patterns. Another example will be a
metalanguage common to many graphical interface users (recall my
analysis of Acrobats interface, which combines metaphors drawn
from diferent media traditions).
Since moving images today systematically combine techniques
of diferent visual media that almost never met until middle of
the 1990s, we are justied in using the term metalanguage in
their case. Visual design today has its own metalanguage, itself is
a subset of the metalanguage of moving images. The reason is that
a designer of moving images has access to all the techniques of a
visual designer plus extra techniques since s/he is working with
additional dimension of time. These two metalanguages also largely
overlap in patterns that are common to thembut there are some
important diferences. For instance, todays moving image works
often feature a continuous movement through a 3D space that may
contain various 2D elements. In contrast, visual designs for print,
web, products, or other applications are usually two-dimensional
they assemble elements over an imaginary at surface. (I think that
the main reason for this insistence on atness is that these designs
often exist next to large blocks of text that already exist in 2D.)
Layers, transparency, compositing
So far I have focused on describing the aesthetics of moving images
that emerged from the Velvet Revolution. While continuing this
investigation, we will now pay more attention to an analysis of the
new software production environment that made this aesthetics
This denition is adopted from Bordwell and Thompson, Film Art, p. 355.
9781623568177_txt_print.indd 277 29/04/2013 11:51
possible. The following sections of this chapter will look at the
tools ofered by After Efects and other media authoring applica-
tions, their user interfaces, and the ways these applications are
used together in production (i.e. design workow). Rather than
discussing all the tools and interface features, I will highlight
a number of fundamental assumptions behind themways of
understanding what a moving image project iswhich, as we will
see, are quite diferent from how it was understood during the
twentieth century.
Probably most dramatic among the changes that took place during
199399 was the new ability to combine multiple levels of imagery
with varying degrees of transparency via digital compositing.

If you
compare a typical music video or a TV advertising spot circa 1986
with its counterpart circa 1996, the diferences are striking. (The
same holds for other areas of visual design.) As I have already noted,
in 1986 computerized memory banks were very limited in their
storage capacity and prohibitively expensive, and therefore designers
could not quickly and easily cut and paste multiple image sources. But
even when they could assemble multiple visual references, a designer
only could place them next to, or on top of each other. S/he could
not modulate these juxtapositions by precisely adjusting transparency
levels of diferent images. Instead, s/he had to resort to the same
photo-collage techniques popularized in the 1920s. In other words,
the lack of transparency restricted the number of diferent image
sources that could be integrated within a single composition without
it starting to look like certain photomontages or photo-collages of
John Hearteld, Hannah Hoch, or Robert Rauschenberga mosaic
of fragments without any strong dominant.
In addition to allowing the superimposition of many transparent
layers, digital compositing also made trivial another operation that
was previously very cumbersome. Until the 1990s, diferent media
types such as hand-drawn animation, lens-based recordings, and
typography practically never appeared within the same frame.
Instead, animated commercials, publicity shorts, industrial lms, and
some feature and experimental lms that did include multiple media
In the case of video, one of the things that made combining multiple visuals
difcult was the rapid degradation of the video signal when an analog video tape
was copied more than a couple of times. Such a copy would no longer meet broad-
casting standards.
9781623568177_txt_print.indd 278 29/04/2013 11:51
usually placed them in separate shots. A few directors have managed
to build whole aesthetic systems out of such temporal juxtaposi-
tionsmost notably, Jean-Luc Godard. In his 1960s lms such
as Weekend (1967) Godard cut bold typographic compositions in
between live action creating what can be called a media montage
(as opposed to a montage of live action shots, as developed by
the Russians in the 1920s). Also in the 1960s, pioneering motion
graphics designer Pablo Ferro, who appropriately called his company
Frame Imagery, created promotional shorts and TV graphics that
played on juxtapositions of diferent media replacing each other in
rapid succession.
In a number of Ferros spots, static images of
diferent letterforms, line drawings, original hand-painted artwork,
photographs, very short clips from newsreels, and other visuals
come one after another with machine gun speed.
Within cinema, the superimposition of diferent media within
the same frame was usually limited to the two media placed on top
of each other in a standardized manneri.e. static letters appearing
on top of still or moving lens-based images in feature lm titles.
In the 1960s, both Ferro and another motion graphics pioneer
Saul Bass created a few remarkable title sequences in which visual
elements of diferent origin were systematically overlaid together
more dynamicallysuch as the opening for Hitchcocks Vertigo
designed by Bass (1958). (Basss 1959 title sequence for North by
Northwest is considered to be the rst to use type in motion).
But I think it is fair to say that such complex juxtapositions of
media within the same frame were rare exceptions in the otherwise
unimedia universe, where lmed images appeared in feature
lms and hand-drawn images appeared in animated lms. The only
twentieth-century feature lm director I know of who has built
his unique aesthetics by systematically combining diferent media
within the same frame was Karel Zeman. Thus, a typical shot by
Zeman may contain lmed human gures, an old engraving used
for background, and a miniature model.

The achievements of these directors and designers are particu-
larly remarkable given the difculty of combining diferent media
Jef Bellantoni and Matt Woolman, Type in Motion (Rizzoli, 1999), pp. 229.
While special efects in feature lms often combined diferent media, they were
used together to create a single illusionistic space, rather than juxtaposed for the
aesthetic efect such as in lms and titles by Godard, Zeman, Ferro, and Bass.
9781623568177_txt_print.indd 279 29/04/2013 11:51
within the same frame during the lm era. To do this required
utilizing the services of special efects departments or separate
companies which used optical printers. Techniques that were cheap
and more accessible, such as double exposure, were limited in their
precision. So while a designer of static images could at least cut
and paste multiple elements within the same composition to create
a photomontage, to produce the equivalent efect with moving
images was far from trivial.
To put this in more general terms, we can say that before the
computerization of the 1990s, the designers capacities to access,
manipulate, remix, and lter visual information, whether still or
moving, were quite restricted. In fact, they were practically the
same as a hundred years earlierregardless of whether lmmakers
and designers used in-camera efects, optical printing, or video
keying. In retrospect, we can see they were at odds with the exi-
bility, speed, and precision of data manipulation that was already
available to most other by then computerized professional elds
sciences, engineering, accounting, management, etc. Therefore it
was only a matter of time before all image media would be turned
into digital data and illustrators, graphic designers, animators, lm
editors, video editors, and motion graphics designers would start to
manipulate them via software instead of their traditional tools. But
this is only obvious todayafter the Velvet Revolution has taken
In 1985 Jef Stein directed a music video for the single You
Might Think by new wave band The Cars. This video was one
of the rst to systematically use computer graphics; it had a big
impact in the design world, and MTV gave it the rst prize in its
rst annual music awards. Stein managed to create a surreal world
in which a video cutout of the singing head of the band member was
animated over diferent video backgrounds. In other words, Stein
took the aesthetics of animated cartoons2D animated characters
superimposed over a 2D backgroundand recreated it using video
imagery. In addition, simple computer animated elements were
also added in some shots to enhance the surreal efect. This was
shocking because no one had ever seen such juxtapositions before.
Suddenly, modernist photomontage came alive. But ten years later,
such moving video collages became not only commonplacethey
also became more complex, more layered, and more subtle. Instead
of two or three, a composition could now feature hundreds and
9781623568177_txt_print.indd 280 29/04/2013 11:51
even thousands of layers. And each layer could have its own level
of transparency.
In short, digital compositing now allowed the designers of
moving images to easily combine any number of visual elements
regardless of the media in which they originated and to control
each element in the process. I can make an analogy between
multitrack audio recording and digital compositing. In multitrack
recording, each sound track can be manipulated individually to
produce the desired result. Similarly, in digital compositing each
visual element can be independently modulated in a variety of
ways: resized, recolored, animated, etc. Just as the music artist can
focus on a particular track while muting all other tracks, a designer
often turns of all visual tracks except the one s/he is currently
adjusting. Similarly, both a music artist and a designer can at any
time substitute one element of a composition by another, delete any
elements, and add new ones. Most importantly, just as multitrack
recording redened the sound of popular music from the 1970s
onward, once digital compositing became widely available during
the 1990s it fundamentally changed the visual aesthetics of most
moving images forms.
This discussion only scratched the surface of my subject
in this section: layers and transparency. For instance, I have
not analyzed the actual techniques of digital compositing and
the fundamental concept of an alpha channel, which deserves
a separate and detailed treatment. I have also not gone into
the possible media histories leading to digital compositing nor
examined its relationship to optical printing, video keying,
and video efects technology of the 1980s. These histories and
relationships were discussed in the Compositing chapter in
The Language of New Media but from a diferent perspective
than the one used here. At that time (1999) I was looking at
compositing from the point of view of the questions of cinematic
realism, practices of montage, and the construction of special
efects in feature lms. Today, however, it is clear to me that in
addition to disrupting the regime of cinematic realism in favor
of other visual aesthetics, compositing also had another, even
more fundamental efect.
By the end of the 1990s digital compositing had become the
basic operation used in creating all forms of moving images, and
not only big budget features. So while it was originally developed
9781623568177_txt_print.indd 281 29/04/2013 11:51
as a technique for special efects in the 1970s and early 1980s,

compositing had a much broader efect on contemporary visual
and media cultures beyond special efects. Compositing played the
key role in turning the digital computer into a kind of experimental
lab (or a Petri dish) where diferent media can meet and where their
aesthetics and techniques can be combined to create new species.
In short, digital compositing was essential in enabling the devel-
opment of a new hybrid visual language of moving imagesone
that today we see everywhere.
Dened at rst as a particular digital technique designed to
integrate the two media of live action lm and computer graphics
in special efects sequences, compositing later became a universal
media integrator. And although compositing was originally
created to support the aesthetics of cinematic realism, over time it
actually had an opposite efect. Rather than forcing diferent media
to fuse seamlessly, compositing led to the ourishing of numerous
media hybrids where the juxtapositions between live and algorith-
mically generated, two dimensional and three dimensional, raster
and vector media are made deliberately visible rather than being
After Eects interface: from time-based
to composition-based
My thesis about media hybridity applies both to the cultural objects
and the software used to create them. Just as the moving image
media made by designers today mix the formats, assumptions, and
techniques of diferent media, the toolboxes and interfaces of the
software they use are also remixes. Let us again use After Efects as
the case study to see how its interface remixes previously distinct
working methods of diferent disciplines.
When moving image designers started to use compositing/
animation software such as After Efects, its interface encouraged
them to think about moving images in a fundamentally new way.
Film and video editing systems and their computer simulations
Thomas Porter and Tom Duf, Compositing Digital Images, ACM Computer
Graphics 18, no. 3 (July 1984): pp. 2539.
9781623568177_txt_print.indd 282 29/04/2013 11:51
that came to be known as non-linear editors (currently exemplied
by Avid, Premiere, and Final Cut
) have conceptualized a media
project as a sequence of shots organized in time. Consequently,
while NLE (the standard abbreviation for non-linear video editing
software) gave the editor many tools for adjusting the edits, they
took for granted the constant of a lm language that came from
its industrial organizationthat all frames have the same size and
aspect ratio. This is an example of a larger trend. During the rst
stage of the development of cultural software, its pioneers were
exploring the new possibilities of a computer metamedium going
in any direction they were interested, since commercial use (with
a notable exception of CAD) was not yet an option. However,
beginning with the 1980s, a new generation of companiesAldus,
Autodesk, Macromedia, Adobe, and othersstarted to produce
GUI-based software media authoring software aimed at particular
industries: TV production, graphic design, animation, etc. As a
result, many of the workow principles, interface conventions and
constraints of media technologies standard in these industries, were
methodically re-created in softwareeven though the software
medium itself has no such limitations. NLE software is a case in
point. In contrast, from the beginning the After Efects interface
put forward a new concept of moving image as a composition
organized both in time and 2D space.
The center of After Efects interface is a Composition concep-
tualized as a large canvas that acts as a background for visual
elements which can have arbitrary sizes, proportions, and content
(video, photos, abstract graphics, type, etc). When I rst started
using After Efects soon after it came out, I remember feeling
shocked that the software did not automatically resize the graphics
I dragged into the Composition window to make them t the
overall frame. The fundamental assumption of cinema that accom-
panied it throughout its whole historythat lm consists of many
frames which all have the same size and aspect ratiowas gone.
In lm and video editing paradigms of the twentieth century, the
minimal unit on which the editor works is a frame. S/he can change
the length of an edit, adjusting where one lm or video segment ends
Compositing functionality was gradually added over time to most NLE systems,
so today the distinction between Efects or Flame interfaces and Avid and Final Cut
interfaces is less pronounced.
9781623568177_txt_print.indd 283 29/04/2013 11:51
The center of After Efects interface is a Composition conceptualized
as a large canvas that acts as a background for visual elements which
can have arbitrary sizes, proportions, and content. Each element can
be individually accessed, manipulated, and animated. The illustration
shows selected frames from a simple 2D animation, as they appear in the
Composition panel.
9781623568177_txt_print.indd 284 29/04/2013 11:51
and another begins, but s/he cannot directly modify the contents of
a frame. The frame functions as a kind of black box that cannot
be opened. (This was the job for special efects departments and
companies.) But in the After Efects interface, the basic unit is not a
frame but a visual element placed in the Composition window. Each
element can be individually accessed, manipulated, and animated.
In other words, each element is conceptualized as an independent
object. Consequently, a media composition is understood as a
set of independent objects that can change over time. The very
word composition is important in this context as it references
2D media (drawing, painting, photography, design) rather than
lmmakingi.e. space as opposed to time.
Where does the After Efects interface come from? Given that
this software is commonly used to create animated graphics and
visual efects, it is not surprising that its interface elements can
be traced to three separate elds: animation, graphic design,
and special efects. And because these elements are integrated in
intricate ways to ofer the user a new experience that cannot be
simply reduced to a sum of the working methods already available
in separate elds, it makes sense to think of the After Efects UI as
an example of deep remixability.
In twentieth-century cel animation production, an animator
places a number of transparent cels on top of each other. Each cel
contains a diferent drawingfor instance, a body of a character on
one cel, the head on another cel, eyes on the third cel. Because the
cels are transparent, the drawings get automatically composited
into a single composition. While the After Efects interface does not
use the metaphor of a stack of transparent cels directly, it is based
on the same principle. Each element in the Composition window is
assigned a virtual depth relative to all other elements. Together
all elements form a virtual stack. At any time, the designer can
change the relative position of an element within the stack, delete
it, or add new elements.
We can also see a connection between the After Efects interface
and another popular twentieth-century animation techniquestop
motion. To create stop motion shots, puppets or any other 3D
objects are positioned in front of a lm camera and manually
animated one step at a time. For instance, an animator may be
adjusting the head of a character, progressively moving it from
left to right in small discrete steps. After every step, the animator
9781623568177_txt_print.indd 285 29/04/2013 11:51
exposes one frame of lm, then makes another adjustment, exposes
another frame, and so on. (The twentieth-century animators and
lmmakers who used this technique with great inventiveness include
Wladyslaw Starewicz, Oskar Fischinger, Aleksandr Ptushko, Ji
Trnka, Jan Svankmajer, and the Brothers Quay.)
Just as in cel and stop-motion animation practices, After Efects
does not make any assumptions about the size or positions of
individual elements. Instead of dealing with standardized units of
timei.e. lm frames containing xed visual contenta designer
now works with separate visual elements. An element can be a
digital video frame, a line of type, an arbitrary geometric shape,
etc. The nished work is the result of a particular arrangement
of these elements in space and time. Consequently, a designer
who uses After Efects can be compared to a choreographer who
creates a dance by animating the bodies of dancersspecifying
their entry and exit points, their trajectories through the space of
the stage, and the movements of their bodies. (In this respect it
is relevant that although the After Efects interface did not evoke
this reference, another equally important 1990s software that was
commonly used to design multimediaMacromedia Directordid
explicitly refer to the metaphor of the theatre stage in its UI.)
While we can link the After Efects interface to traditional
animation methods as used by commercial animation studios,
the working method put forward by software is closer to graphic
design. In commercial animation studios of the twentieth century
all elementsdrawings, sets, characters, etc.were prepared
beforehand. The lming itself was a mechanical process. Of
course, we can nd exceptions to this industrial-like separation of
labor in experimental animation practice where a lm was usually
produced by one person. This allowed a lmmaker to invent
a lm as he went along, rather than having to plan everything
beforehand. A classic example of this is Oskar Fischingers Motion
Painting 1 created in 1949. Fischinger made this 11-minute lm
by continuously modifying a painting and exposing lm one frame
at a time after each modication. This process took nine months.
Because Fischinger was shooting on lm, he had to wait a long time
before seeing the results of his work. As the historian of abstract
animation William Moritz writes, Fischinger painted every day
for over ve months without being able to see how it was coming
out on lm, since he wanted to keep all the conditions, including
9781623568177_txt_print.indd 286 29/04/2013 11:51
lm stock, absolutely consistent in order to avoid unexpected
variations in quality of image.
In other words, in the case of this
project by Fischinger, creating animation and seeing the result were
even more separated than in a commercial animation process.
In contrast, a graphic designer works in real time. As the
designer introduces new elements, adjusts their locations, colors
and other properties, tries diferent images, changes the size of
the type, and so on, s/he can immediately see the result of his/her
After Efects adopts this working method by making the
Composition window the center of its interface. Like a traditional
designer, the After Efects user interactively arranges the elements in
this window and can immediately see the result. In short, the After
Efects interface makes lmmaking into a design process, and a lm
is re-conceptualized as a graphic design that can change over time.
As we saw when we looked at the history of cultural software,
when physical or electronic media are simulated in a computer, we
do not simply end up with the same media as before. By adding
new properties and working methods, computer simulation funda-
mentally changes the identity of a given media. For example, in
the case of electronic paper such as a Word document or a PDF
le, we can do many things which were not possible with ordinary
paper: zoom in and out of the document, search for a particular
phrase, change fonts and line spacing, etc. Similarly, online inter-
active maps services provided by Google and Microsoft augment
the traditional paper map in multiple and amazing ways.
A signicant proportion of contemporary software for creating,
editing, and interacting with media was developed in this way.
Quoted in Michael Barrier, Oskar Fischinger. Motion Painting No. 1, http://www.
Depending on the complexity of the project and the hardware conguration,
the computer may or may not be able to keep pace with the designers changes.
Often a designer has to wait until the computer renders everything in frame after
s/he makes changes. However, since s/he has control over this rendering process,
s/he can instruct After Efects to show only outlines of the objects, to skip some
layers, etc.thus giving the computer less information to process and allowing for
real-time feedback. While a graphic designer does not have to wait until a lm is
developed or a computer has nished rendering the animation, the design has its
own rendering stagemaking proofs. With both digital and ofset printing, after
the design is nished, it is sent to the printer who produces the test prints. If the
designer nds any problems such as incorrect colors, s/he adjusts the design and
then asks for proofs again.
9781623568177_txt_print.indd 287 29/04/2013 11:51
Already existing media technologies were simulated in a computer
and augmented with new properties. But if we consider media
authoring software such as Maya (3D modeling and computer
animation) or After Efects (motion graphics, compositing and
visual efects), we encounter a diferent logic. These software
applications do not simulate any single physical media that existed
previously. Rather, they borrow from a number of diferent
media, combining and mixing their working methods and specic
techniques. (And, of course, they also add new capabilities specic
to computersfor instance, the ability to automatically calculate
the intermediate values between a number of key frames.) For
example, 3D modeling software mixes form-making techniques
which were previously hardwired to diferent physical media:
the ability to change the curvature of a rounded form as though it
were made from clay, the ability to build a complex 3D object from
simple geometric primitives the way buildings were constructed
from identical rectangular bricks, cylindrical columns, pillars, etc.
Similarly, as we saw, the After Efects original interface, toolkit,
and workow draws upon the techniques of animation and graphic
design. (We can also nd traces of lmmaking and 3D computer
graphics.) But the result is not simply a mechanical sum of all
elements that came from earlier media. Rather, as software remixes
the techniques and working methods of the various media they
simulate, the results are new interfaces, tools, and workows with
their own distinct logic. In the case of After Efects, the working
method that it puts forward is neither animation nor graphic
design nor cinematography, even though it draws from all these
elds. It is a new way to make moving image media. Similarly, the
visual language of media produced with this and similar software
is also diferent from the languages of moving images that existed
Consequently, the Velvet Revolution unleashed by After Efects
and other software did not simply make more commonplace the
animated graphics that artists and designers such as John and James
Whitney, Norman McLaren, Saul Bass, Robert Abel, Harry Marks,
Richard and Robert Greenberg were previously creating using stop
motion animation, optical printing, video efects hardware of the
1980s, and other custom techniques and technologies. Instead it
led to the emergence of numerous new visual aesthetics that did
not exist before. And if the common feature of these aesthetics is
9781623568177_txt_print.indd 288 29/04/2013 11:51
deep remixability, it is not hard to see that it mirrors the deep
remixability in the After Efects UI.
3D space as a media design platform
As I was researching what users and industry reviewers had been
saying about After Efects, I came across a somewhat condescending
characterization of this software as Photoshop with keyframes.
I think that this characterization is actually quite useful.
about all the diferent ways of manipulating images available in
Photoshop and the degree of control provided by its multiple tools.
Think also about Photoshops concept of a visual composition
as a stack of potentially hundreds of layers, each with its own
transparency setting and multiple alpha channels. If we are able to
animate such a composition and continue using Photoshops tools
to adjust visual elements over time on all layers independently, this
indeed constitutes a new paradigm for creating moving images.
And this is what After Efects and other animation, visual efects,
and compositing software make possible today.
And while the
idea of working with a number of layers placed on top of each
other itself is not newconsider traditional cel animation, optical
printing, video switchers, photo-collage, graphic designgoing
from a few non-transparent layers to hundreds and even thousands
of layers, each with its controls, fundamentally changes not only
how a moving image looks but also what it can say. From being a
special efect reserved for particular shots, 2D compositing became
a part of the standard animation and video editing interface.
But innovative as the 2D compositing paradigm was, by the
beginning of the 2000s it was supplemented by a new one: 3D
compositing. If 2D compositing can be thought as an extension
Soon after the initial release of After Efects in January 1993, CoSA (the company
that produced this software) was purchased by Aldus, which in turn was purchased
by Adobewhich was already selling Photoshop.
Photoshop and After Efects were designed originally by diferent teams at
diferent times, and even after both products were purchased by Adobe (it released
Photoshop in 1989 and After Efects in 1995), it took Adobe a number of years to
build close links between After Efects and Photoshop eventually making it easy to
move assets between the two programs.
9781623568177_txt_print.indd 289 29/04/2013 11:51
of already familiar media techniques, the new paradigm does not
come from any previous physical or electronic media. Instead, it
takes the new born-digital media which was invented in the 1960s
and matured by the early 1990sinteractive 3D computer graphics
and animationand transforms it into a general platform for
moving media design.
The language used in the professional production milieu
today reects an implicit understanding that 3D graphics is a
new medium unique to a computer. When people use the terms
computer visuals, computer imagery, or CGI (which is
an abbreviation for computer generated imagery) everybody
understands that they refer to 3D graphics as opposed to other
image sources such as digital photography. I think of 3D
computer graphics as a new mediaas opposed to considering
it an extension of architectural drafting, projection geometry, or
set makingbecause they ofer a new method for representing
three-dimensional reality: both objects that already exist and
objects that are only imagined. This method is fundamentally
diferent from what has been ofered by the main representational
media of the industrial era: lens-based capture (still photography,
lm recording, video) and audio recording. With 3D computer
graphics, we can represent a three-dimensional structure of the
world, versus capturing only a perspectival image of the world,
as in lens-based recording. We can also manipulate our represen-
tation using various tools with ease and precisionqualitatively
diferent from the much more limited manipulability of a model
made from any physical material (although nanotechnology
promises to change this in the future). And, as contemporary
architectural aesthetics makes clear, 3D computer graphics is
not simply a faster way of working with geometric representa-
tions like the plans and cross-sections used by draftsmen for
centuries. When generations of young architects and architec-
tural students started to systematically work with 3D modeling
and animation software such as Alias in the middle of the 1990s,
the ability to directly manipulate a 3D shape (rather than only
dealing with its projections as in traditional drafting) quickly
led to a whole new language of complex non-rectangular curved
forms. In other words, architects working with the media of 3D
computer graphics started to imagine diferent things than their
predecessors who used pencils, rules, and drafting tables.
9781623568177_txt_print.indd 290 29/04/2013 11:51
When the Velvet Revolution of the 1990s made it possible to
easily combine multiple media sources in a single moving image
sequence using multi-layer interface of After Efects, CGI was
added to the mix. Today, 3D models are routinely used in media
compositions created in After Efects and similar software, along
with all other media sources. But in order to be a part of the mix,
these models need to be placed on their own 2D layers and thus
treated as 2D images. This was the original After Efects paradigm:
all image media can meet as long as they are reduced to 2D.
In contrast, in the 3D compositing paradigm all media types
are placed within a single 3D space. One advantage of this
representation is that since 3D space is native to 3D computer
graphics, 3D models can stay as they are, i.e. three-dimensional.
An additional advantage is that the designer can now use all the
techniques of virtual cinematography as developed in 3D computer
animation. S/he can dene diferent kinds of lights, y the virtual
camera around and through the image planes using any trajectory,
and use depth of eld and motion blur efects.

While 3D computer-generated models already live in this
space, how do you bring in two-dimensional visual elements
video, digitized lm, typography, drawn images? If 2D compositing
paradigm treated everything as 2D imagesincluding 3D computer
models3D compositing treats everything as 3D. So while
two-dimension elements do not inherently have a third dimension,
it has to be added to enable these elements to enter the three-
dimensional space. To do that, a designer places at cards in this
I say original because the later version of After Efects added the ability to work
with 3D layers.
If 2D compositing can be understood as an extension of twentieth-century cel
animation, where a composition consists of a stack of at drawings, the conceptual
source of the 3D compositing paradigm is diferent. It comes out of the work on
integrating live action footage and CGI done in the 1980s in the context of feature
lms production. Both lm director and computer animator work in a three-dimen-
sional space: the physical space of the set in the rst case, the virtual space as dened
by 3D modeling software in the second case. Therefore conceptually it makes sense
to use three-dimensional space as a common platform for the integration of these
two worlds. It is not accidental that NUKE, one of todays leading programs for 3D
compositing was developed in-house at Digital Domain, which was co-founded in
1993 by James Cameron, the Hollywood director who has systematically advanced
the integration of CGI and live action in lms such as The Abyss (1989), Terminator
2 (1991), and Titanic (1997).
9781623568177_txt_print.indd 291 29/04/2013 11:51
In the 3D compositing paradigm all media types are placed within a
single 3D space. This illustration uses the animation from earlier, buts
adds the third dimension. We positioned each element at a diferent
depth, and added a light, casting shadows.
9781623568177_txt_print.indd 292 29/04/2013 11:51
space in particular locations, and situates two-dimensional images
on these cards. Now, everything lives in a common 3D space. This
condition enables deep remixability between techniques which I
have illustrated using the example of Go video. The techniques
of drawing, photography, cinematography, and typography which
go into capturing or creating two-dimensional visual elements
can now play together with all the techniques of 3D computer
animation (virtual camera moves, controllable depth of eld,
variable lens, etc.)
In 1995 I wrote the article What is Digital Cinema? The
article asked how the changes in moving image production I was
witnessing (I was living in Los Angeles at that time, which made it
easy to follow what was happening in Hollywood) were afecting
the meaning of cinema. In that article I proposed that the logic
of hand-drawn animation, which throughout the twentieth century
was marginal in relation to cinema, became dominant in a software
era. Because software allows the designer to manually manipulate
any image regardless of its source as though it was drawn in the
rst place, the ontological diferences between diferent image
media become irrelevant. Both conceptually and practically, they
are all reduced to hand-drawn animation.
By default, After Efects and other animation/video editing/2D
compositing software treats a moving image project as a stack of
layers. Therefore, I can extend my original argument and propose
that animation logic also moves from the marginal to the dominant
position in yet another way. The paradigm of a composition as
a stack of separate visual elements as practiced in cel animation
becomes the default way of working with all images in a software
environmentregardless of their origin and nal output media. In
other words, a moving image is now understood as a composite
of layers of imageryrather than as a still at picture that only
changes in time, as it was the case for most of the twentieth century.
In the word of animation, editing, and compositing software, such
a single layer image becomes an exception.
The emergence of the 3D compositing paradigm can be also seen
as following this logic of historical reversal. The new representa-
tional structure as developed within the computer graphics elda
3D virtual space containing 3D modelshas gradually moved from
a marginal to a dominant role. In the 1970s and 1980s computer
graphics were used only occasionally in a dozen feature lms such
9781623568177_txt_print.indd 293 29/04/2013 11:51
as Alien (1979), Tron (1981), The Last Starghter (1984), and The
Abyss (1989), and selected television commercials and broadcast
graphics. But by the beginning of the 2000s, the representation
structure of computer graphics, i.e. a 3D virtual space, came to
function as an umbrella for all other image types regardless of
their origin. An example of an application which implements this
paradigm is Flame, enthusiastically described by one user as a
full 3D compositing environment into which you can bring 3D
models, create true 3D text and 3D particles, and distort layers in
3D space.

This does not mean that 3D animation itself became visually
dominant in moving image culture, or that the 3D structure of
the space within which media compositions are now routinely
constructed is necessarily made visible (usually it is not). Rather,
the way 3D computer animation organizes visual dataas
objects positioned in a Cartesian spacebecame the way to
work with all moving image media. As already stated above, a
designer positions all the elements which go into a composition
2D animated sequences, 3D objects, particle systems, video and
digitized lm sequences, still images and photographsinside
the shared 3D virtual space. There these elements can be further
animated, transformed, blurred, ltered, etc. So while all moving
image media has been reduced to the status of hand-drawn
animation in terms of their manipulability, we can also state
that all media have become layers in 3D space. In short, the new
media of 3D computer animation has eaten up the dominant
media of the industrial agelens-based photo, lm and video
Having discussed how software has redened the concept of
a moving image as a composite of multiple layers, this is a
good moment to pause and consider other possible ways software
changed this concept. When cinema in its modern form was born in
the end of the nineteenth century, the new medium was understood
as an extension of already familiar onethat is, as a photographic
image which was now moving. This understanding can be found in
the press accounts of the day and also in one of the ofcial names
given to the new mediummoving pictures. On the material
Alan Okey, post to forums.creativecow.net, December 28, 2005, http://forums.
9781623568177_txt_print.indd 294 29/04/2013 11:51
level, a lm indeed consisted of separate photographic frames.
When they quickly replace each other, this creates the efect of
motion for the viewer. So the concept used to understand cinema
indeed t in with the structure of the medium.
But is this concept still appropriate today? When we record
video and play it, we are still dealing with the same structure: a
sequence of frames. But for professional media designers, the terms
have changed. The importance of these changes is not just academic
and purely theoreticalbecause designers understand their media
diferently, they are creating lms and sequences that also look very
diferent from twentieth-century cinema or animation.
Consider those new ways of creating moving images (which I
referred to as a new paradigms) that we have discussed thus far.
(Although theoretically they are not necessarily all compatible
with each other, in production practice these diferent paradigms
are used in a complementary fashion.) A moving image became
a hybrid which can combine all diferent visual media invented
so farrather than holding only one kind of data such as camera
recording, hand drawing, etc. Rather than being understood as
a singular at planethe result of light focused by the lens and
captured by the recording surfaceit is now understood as a
stack of a potentially innite number of separate layers. And
rather than time-based, it becomes composition-based, or
objectoriented. That is, instead of being treated as a sequence
of frames arranged in time, a moving image is now understood
as a two-dimensional composition that consists of a number of
objects that can be manipulated independently. Alternatively, if a
designer uses 3D compositing, the conceptual shift is even more
dramatic: instead of editing images, s/he is working in a virtual
three-dimensional space that holds both CGI and lens-recorded at
image sources.
Of course, frame-based representation did not disappearit
became simply a recoding and output format rather than the space
where a lm is being put together. And while the term moving
image can be still used as an appropriate description for how the
output of a production process is experienced by the viewers, it is
no longer captures how the designers think about what they create.
Because their production environmentworkow, interfaces, and
toolshas changed so much, they are thinking today very difer-
ently than twenty years ago.
9781623568177_txt_print.indd 295 29/04/2013 11:51
If we focus on what the diferent paradigms summarized above
have in common, we can say that lmmakers, editors, special efects
artists, animators, and motion graphics designers are all working
on a composition in 2D or a 3D space that consists of a number of
separate objects. The spatial dimension has become as important
as the temporal dimension. From the concept of a moving image
understood as a sequence of static photographs we have moved
to a new concept: a modular media composition. And while a
person who directs a feature or a short lm that is centered around
actors and live action can be still called a lmmaker, in all other
cases where most of the production takes place in a software
environment, it is more appropriate to call the person a designer.
This is yet another fundamental change in the concept of moving
images: today more often than not they are not captured,
directed, or animated. Instead, they are designed.
Import/export: design workow
In our discussions of the After Efects interface and workow, as
well as the newer paradigm of 3D compositing, we have already
come across the crucial aspect of software-based media production
process. Until the arrival of software-based tools in the 1990s,
combining diferent types of time-based media together was either
time consuming, expensive, or in some cases simply impossible.
Software tools such as After Efects have changed this situation in a
fundamental way. Now a designer can import diferent media into
their composition with just a few mouse clicks.
However, the contemporary software-based design of moving
imagesor any other design process, for that matterdoes not
simply involve combining elements from diferent sources within
a single application. In this section we will look at the whole
workow typical of contemporary designbe it the design of
moving images, still illustrations, 3D objects and scenes, archi-
tecture, music, websites, or any other media. (Most of the details
of software-based production of moving images which I have
already presented also applies to the graphic design of still images
and layouts for print, the web, packaging, physical spaces, mobile
devices, etc. However, in this section I want to make this explicit.
9781623568177_txt_print.indd 296 29/04/2013 11:51
Therefore the examples below will include not only moving images,
but also graphic design.)
Although import/export commands appear in most modern
media authoring and editing software applications running under a
GUI, at rst sight they do not seem to be very important for under-
standing software culture. When you import, you are neither
authoring new media nor modifying media objects, nor accessing
information across the globe, as in web browsing. All these commands
allow you to do is to move data around between diferent applications.
In other words, they make data created in one application compatible
with other applications. And that does not look so important.
Think again. What is the largest part of the economy of the
greater Los Angeles area? It is not entertainment. From movie
production to museums and everything in between, entertainment
only accounts for 15 percent. It turns out that the largest part
of the economy is the import/export businessmore than 60
percent. More generally, one commonly evoked characteristic of
globalization is greater connectivityplaces, systems, countries,
organizations etc. becoming connected in more and more ways. And
connectivity can only happen if you have certain level of compat-
ibility: between business codes and procedures, between shipping
technologies, between network protocols, between computer le
formats, and so on.
Let us take a closer look at import/export commands. As I will
try to show below, these commands play a crucial role in software
culture, and in particular in media designregardless of what kind
of project a design is working on.
Before they adopted software tools in the 1990s, lmmakers,
graphic designers, and animators used completely diferent
technologies. Therefore, as much as they were inuenced by each
other or shared the same aesthetic sensibilities, they inevitably
created diferent-looking images. Filmmakers used camera and
lm technology designed to capture three-dimensional physical
reality. Graphic designers were working with ofset printing and
lithography. Animators were working with their own technologies:
transparent cels and an animation stand with a stationary lm
camera capable of making exposures one frame at a time as the
animator changed cels and/or moved background.
As a result, twentieth-century cinema, graphic design, and
animation (I am talking here about standard animation techniques
9781623568177_txt_print.indd 297 29/04/2013 11:51
used by most commercial studios) developed distinct artistic
languages and vocabularies both in terms of form and content. For
example, graphic designers worked with a two dimensional space,
lm directors arranged compositions in three-dimensional space,
and cel animators worked with two-and-a-half dimensions.
This holds for the overwhelming majority of works produced in
each eld, although of course exceptions do exist. For instance,
Oskar Fischinger made one abstract lm that consisted of simple
geometric objects moving in an empty spacebut as far as I know
this is the only lm in the history of abstract pre-digital animation
that takes place in three-dimensional space.
Diferences in technology inuenced what kind of content
would appear in diferent media. Cinema showed photoreal-
istic images of nature, built environments and human forms
articulated by special lighting. Graphic designs featured typog-
raphy, abstract graphic elements, monochrome backgrounds, and
cutout photographs. And cartoons presented hand-drawn at
characters and objects animated over hand-drawn (but more
detailed) backgrounds. The exceptions are rare. For instance, while
architectural spaces frequently appeared in lms because directors
could explore their threedimensionality in staging scenes, they
practically never appeared in animated lms in any detailuntil
animation studios started using 3D computer animation.
Why was it so difcult to cross boundaries? For instance, in
theory one could imagine making an animated lm in the following
way: printing a series of slightly diferent graphics designs and then
lming them as though they were a sequence of animated cels. Or
a lm where a designer simply made a series of hand drawings that
used the exact vocabulary of graphic design and then lmed them
one by one. And yet, to the best of my knowledge, such a lm was
never made. What we nd instead are many abstract animated
lms that have a certain connection to various styles of abstract
painting. For example, Oskar Fischingers lms and paintings
share certain forms. We can also nd abstract lms and animated
commercials and movie titles that have a connection to graphic
design aesthetics popular around the same time. For instance, some
moving image sequences made by motion graphics pioneer Pablo
Ferro in the 1960s display psychedelic aesthetics which can also be
found in posters, record covers, and other works of graphic design
in the same period.
9781623568177_txt_print.indd 298 29/04/2013 11:51
And yet despite these connections, works in diferent media
never used exactly the same visual language. One reason is that
projected lm could not adequately show the subtle diferences
between typeface sizes, line widths, and gray-scale tones crucial for
modern graphic design. Therefore, when the artists were working
on abstract art lms or commercials that adopted design aesthetics
(and most major twentieth-century abstract animators worked both
on their personal lms and commercials), they could not simply
expand the language of a printed page into the time dimension.
They essentially had to invent a parallel visual language that used
bold contrasts, more easily readable forms, and thick linesthat,
because of their thickness, were in fact no longer lines but shapes.
Although the limitations in resolution and contrast of lm and
television image in comparison to a printed page contributed to the
distance between the languages used by abstract lmmakers and
graphic designers for the most of the twentieth century, ultimately
I do not think it was the decisive factor. Today the resolution,
contrast and color reproduction between print, computer screens,
television screens, and the screens of mobile phones are also
substantially diferentand yet we often see exactly the same
visual strategies deployed across these diferent display media. If
you want to be convinced, leaf through any book or a magazine on
contemporary 2D design (i.e., graphic design for print, broadcast,
and the web). When you look at pages featuring the works of a
particular designer or a design studio, in most cases it is impossible
to identify the origins of the images unless you read the captions.
Only then do you nd which image is a poster, which one is a still
from a music video, and which one is a magazine editorial.
I am going to use Graphic Design for the 21st Century: 100 of the
Worlds Best Graphic Designers (Taschen, 2001) for my examples
because by 2001, the changes I describe had already taken place.
Peter Andersons design showing a line of type against a cloud of
hundreds of little letters in various orientations turns out to be the
frames from the title sequence for a Channel 4 documentary. His
other design which similarly plays on the contrast between jumping
letters in a larger font against irregularly cut planes made from
densely packed letters in much smaller fonts, turns to be a spread
from IT Magazine. Since the rst design was made for broadcast
while the second was made for print, we would expect that the
rst design would employ bolder formshowever, both designs
9781623568177_txt_print.indd 299 29/04/2013 11:51
use the same scale between big and small fonts, and feature texture
elds composed from hundreds of words in such a small font that
they are clearly not intended to be read. A few pages later we
encounter a design by Philippe Apeloig that uses exactly the same
technique and aesthetics as Anderson used. In this case, tiny lines
of text positioned at diferent angles form a 3D shape oating in
space. On the next page another design by Apeloig creates a eld in
perspectivemade not from letters but from hundreds of identical
abstract shapes.
These designs rely on the softwares ability (or on the designer
being inuenced by software use and recreating what s/he did
with software manually) to treat text as any graphical primitive
and to easily create compositions made from hundreds of similar
or identical elements positioned according to some pattern. And
since an algorithm can easily modify each element in the pattern,
changing its position, size, color, etc., instead of the completely
regular grids of modernism we see more complex structures that
are made from many variations of the same element. This strategy
is explored particularly imaginatively in Zaha Hadids designs such
as the Louis Vuitton Icone Bag 2006, and in her urban master
plans for Singapore and Turkey, which use what Hadid called a
variable grid.
Each designer included in the book was asked to provide a
brief statement to accompany the portfolio of his/her work, and
Lust Studio has chosen the phrase Form-follows-process as
their motto. So what is the nature of the design process in the
software age and how does it inuence the forms we see today
around us?
If you are practically involved in design or art today, you
already know that contemporary designers use the same small
set of software tools to design just about everything. I have
already named them repeatedly, so you know the list: Photoshop,
Illustrator, Flash, Maya, etc. However, the crucial factor is not the
tools themselves but the workow process, enabled by import
and export operations and related methods (place, insert
object, subscribe, smart object, etc.), that ensure coordi-
nation between these tools.
When a particular media project is being put together, the
software used at the nal stage depends on the type of output media
and the nature of the projectAfter Efects for motion graphics
9781623568177_txt_print.indd 300 29/04/2013 11:51
Examples of a variable grid in architecture: two master plans by
Zaha Hadid. Top: Kartal Pendik, Istanbul, 2006. Bottom: One North,
Singapore, 2001.
9781623568177_txt_print.indd 301 29/04/2013 11:51
projects and video compositing, Illustrator for print illustrations,
InDesign for multi-page designs, Flash for interactive interfaces and
web animations, 3ds Max or Maya for 3D computer models and
animations, and so on. But these programs are rarely used alone
to create a media design from start to nish. Typically, a designer
may create elements in one program, import them into another
program, add elements created in yet another program, and so on.
This happens regardless whether the nal product is an illustration
for print, a website, or a motion graphics sequence, whether it is a
still or a moving image, interactive or non-interactive, etc.
The very names which software companies give to the products
for media design and production refer to this dening character-
istic of software-based design process. Since 2005, Adobe has been
selling its diferent media authoring applications bundled together
under the name Adobe Creative Suite. Among the subheadings
and phrases that were used to accompany this brand name, one
in particular is highly meaningful in the context of our discussion:
Design Across Media. This phrase accurately describes both the
capabilities of the applications collected in a suite and their actual
use in the real world. Each of the key applications collected in the
suitePhotoshop, Illustrator, InDesign, Flash, Dreamweaver, After
Efects, Premierehas many special features geared for producing
a design for particular output media. Illustrator is set up to work
with professional-quality printers; After Efects and Premiere can
output video les in a variety of standard video formats such
as HDTV; Dreamweaver supports programming and scripting
languages to enable creation of sophisticated and large-scale
dynamic websites. But while a design project is nished in one of
these applications, most other applications in Adobe Creative Suite
will be used in the process to create and edit its various elements.
This is one of the ways in which Adobe Creative Suite enables
design across media. The compatibility between applications also
means that the elements (called in professional language assets)
can be later re-used in new projects. For instance, a photograph
edited in Photoshop can rst be used in a magazine ad and later
put in a video, a website, etc. Or, the 3D models and characters
created for a feature lm can be reused for a video game based
on the lm. This ability to re-use the same design elements for
very diferent projects is very important because of the widespread
practice in creative industries to create products across the range
9781623568177_txt_print.indd 302 29/04/2013 11:51
of media which share the same images, designs, characters, narra-
tives, etc. An advertising campaign often works across media
including web ads, TV ads, magazine ads, billboards, etc. And if
turning movies into games and games into movies has been already
popular in Hollywood since the early 1990s, a new trend since
approximately the middle of the 2000s is to create a movie, a game,
a website or maybe other media products at the same timeand
have all the products use the same digital assets both for economic
reasons and to assure aesthetic continuity between these products.
Thus, a studio may create 3D backgrounds and characters and
put them in both a movie and a game, both of which will be
released simultaneously. If media authoring applications were not
compatible, such a practice would simply not be possible.
All these examples illustrate the intentional reuse of design
elements across media. However, the compatibility between
media authoring applications also has a much broader and uninten-
tional efect on contemporary aesthetics. Given the production
workow I have just described, we may expect that the same visual
techniques and strategies will also appear in all types of media
projects designed with software, without this being consciously
planned for. We may also expect that this will happen on a much
more basic level. This is indeed the case. The same software-
enabled design strategies, the same software-based techniques, and
the same software-generated iconography are now found across all
types of media, all scales, and all kinds of projects.
We have already encountered a few concrete examples. For
instance, the three designs by Peter Anderson and Philippe Apeloig
done for diferent media use the same basic computer graphic
technique: automatic generation of a repeating pattern while
varying the parameters that control the appearance of each element
making up the patternits size, position, orientation, curvature,
etc. (The general principle behind this technique can also be used
to generate 3D models, animations, textures, make plants and
landscapes, etc. It is often referred to as parametric design,
or parametric modeling.) The same technique is also used by
Hadids studio for the Louis Vuitton Icone Bag. In another
example, which will be discussed below, Greg Lynn used particle
systems techniquewhich at that time was normally used to
simulate re, snow, waterfalls, and other natural phenomena in
cinemato generate the forms of a building.
9781623568177_txt_print.indd 303 29/04/2013 11:51
To use the biological metaphor, we can say that compatibility
between design applications creates very favorable conditions for
the propagation of media DNA between species, families, and
classes. And this propagation happens on all levels: the whole
design, parts of a design, the elements making up the parts, and
the atoms that make up the elements. Consider the following
hypothetical example of propagation on a lower level. A designer
can use Illustrator to create a 2D smooth curve (called in the
computer graphics eld a spline.) This curve becomes a building
block that can be used in any project. It can form part of an illus-
tration or a book design. It can be imported into an animation
program where it can be set to motion, or imported into 3D
program where it can be extruded in 3D space to dene a solid
Over time software manufacturers worked to develop tighter
ways of connecting their applications, in order to make it easier
to move elements from one application to another. Over the years,
it became possible to move a whole project between applications
without losing anything (or almost anything). For example, in
describing the integration between Illustrator CS3 and Photoshop
CS3, Adobes website states that a designer can Preserve layers,
layer comps, transparency, and editable les when moving les
between Photoshop and Illustrator.
Another important devel-
opment has been the concept that Microsoft Ofce calls linked
objects. If you link all or part of one le to another le (for
instance, linking an Excel document to a PowerPoint presentation),
anytime information changes in the rst le it automatically gets
updated in the second le. Many media applications implement
this feature. To use the same example of Illustrator CS3, a designer
can Import Illustrator les into Adobe Premiere Pro software,
and then use Edit Original command to open the artwork in
Illustrator, edit it, and see your changes automatically incorpo-
rated into your video project.
Each type of program used by media designers3D graphics,
vector drawing, image editing, animation, compositingexcel
at particular design operations, i.e. particular ways of creating
design elements or modifying already existing elements. These
http://www.adobe.com/products/illustrator/features/allfeatures/ (August 30, 2008).
9781623568177_txt_print.indd 304 29/04/2013 11:51
operations can be compared to the diferent types of blocks of a
LEGO set. You can create an innite number of projects by using
the limited number of block types provided in the set. Depending
on the project, these block types will have diferent functions and
appear in diferent combinations. For example, a rectangular
block may become part of the tabletop, part of the head of a robot,
A design workow that uses a small number of compatible
software programs works in a similar waywith one important
diference. The building blocks used in contemporary design are
not only diferent kinds of visual elements one can createvector
patterns, 3D objects, particle systems, etc.but also various
ways of modifying these elements: blur, skew, vectorize, change
transparency level, extrude, etc. This diference is crucial. If
media creation and editing software did not include these and
many other modication operations, we would see an altogether
diferent visual language at work today. We would see multi-
media, i.e. designs that simply combine elements from diferent
media. Instead, we see deep remixabilitythe deep interac-
tions between working methods and techniques of diferent media
within a single project.
In a crossover use, the techniques which were previously
specic to one media are applied to other media types (for
example, a lens blur lter). This often can be done within a
single applicationfor instance, applying After Efects blur
lter to a composition which can contain graphic elements,
video, 3D objects, etc. And being able to move a whole project
or its elements between applications opens more possibilities
because each application ofers many unique techniques not
available in other applications. As the media data travels from
one application to the next, it is being transformed and enhanced
using the operations ofered by each application. For example,
a designer can take the project s/he has been editing in Adobe
Premiere and import it in After Efects where s/he can use
advanced compositing features of that program. S/he can then
import the result back into Premiere and continue editing. Or s/
he can create artwork in Photoshop or Illustrator and import it
into Flash where it can be animated. This animation can be then
imported into a video editing program and combined with video.
And so on.
9781623568177_txt_print.indd 305 29/04/2013 11:51
The production workow specic to the software era that I have
just illustrated has two major consequences. Its rst result is the
visual aesthetics of hybridity that dominates contemporary design
universe. The second is the use of the same techniques and strat-
egies across this universeregardless of the output media and type
of project.
As I have already stated more than once, a typical design
today combines techniques coming from multiple media. We are
now in a better position to understand why this is the case. As a
designer works on a project, s/he combines the results of the opera-
tions specic to diferent software programs that were originally
created to imitate work with diferent physical media (Illustrator
was created to make illustrations, Photoshop to edit digitized
photographs; Premiere to edit video, etc.) While these operations
continue to be used in relation to their original media, most of
them are now also used as part of the workow on any design job.
The essential condition that enables this new design logic and
the resulting aesthetics is compatibility between les generated
by diferent programs. In other words, import, export and
related functions and commands of graphics, animation, video
editing, compositing and modeling software are historically more
important than the individual operations these programs ofer. The
ability to combine raster and vector layers within the same image,
place 3D elements into a 2D composition and vice versa, etc. is
what enables the production workow and its reuse of the same
techniques, efects, and iconography across diferent media.
The consequences of this compatibility between software and le
formats, which was gradually achieved during the 1990s, are hard
to overestimate. Besides the hybridity of modern visual aesthetics
and the reappearance of exactly the same design techniques
across all output media, there are also other efects. For instance,
the whole eld of motion graphics as it exists today came into
existence to a large extent because of the integration between vector
drawing software, specically Illustrator, and animation/compos-
iting software such as After Efects. A designer typically denes
various composition elements in Illustrator and then imports them
into After Efects where they are animated. This compatibility did
not exist when the initial versions of diferent media authoring and
editing software became available in the 1980s. It was gradually
added in particular software releases. But when it was achieved
9781623568177_txt_print.indd 306 29/04/2013 11:51
around the middle of the 1990s,
within a few years the whole
language of contemporary graphic design was fully imported into
the moving image areaboth literally and metaphorically.
In summary, the compatibility between graphic design, illus-
tration, animation, video editing, 3D modeling and animation, and
visual efects software plays the key role in shaping the visual and
spatial forms of the software age. On the one hand, never before
have we witnessed such a variety of forms as today. On the other
hand, exactly the same techniques, compositions and iconography
can now appear in any media.
Variable form
As the lms of Blake and Murata discussed earlier illustrate, in
contemporary motion graphics the transformations often afect
the frame as a whole. In contrast to twentieth-century animation,
everything inside the frame keeps changing: visual elements, their
transparency, the texture of the image, etc. In fact, if something
stays the same for a while, that is an exception rather than the
Such constant change on many visual dimensions is another key
feature of motion graphics and design cinema produced today.
Just as we did in the case of media hybridity, we can connect this
preference for constant change to the particulars of software used
in media design.
Digital computers allow us to represent any phenomenon or
structure as a set of variables. In the case of design and animation
software, this means that all possible formsvisual, temporal,
spatial, interactiveare similarly represented as sets of variables
that can change continuously. This new logic of form is deeply
encoded in the interfaces of software packages and the tools they
Consider again the After Efects interface. To create an animation,
a designer adds a number of elements to the composition and then
animates them. Each new element shows up in the interface as a
In 1995, After Efects 3.0 enabled importing Illustrator les and Photoshop les
as compositions. http://en.wikipedia.org/wiki/Adobe_After_Effects
9781623568177_txt_print.indd 307 29/04/2013 11:51
To create an animation, a designer adds a number of elements to the
composition and then animates them. Each new element shows up in
the interface as a list of parameters. Animating any parameter is equally
easy, and it only takes a few clicks. The illustration shows a part of
After Efects Timeline panel for the 3D animation illustrated earlier. The
Timeline contains one light, one camera, three shapes (two rectangles and
a circle), and a horizontal plane.
9781623568177_txt_print.indd 308 29/04/2013 11:51
The parameters for the light, the camera, and the circle shape as they
appear in After Efects interface. Each parameter can be animated
separately. (To create this illustration, we redrew a part of the After
Efects Timeline panel screenshot that appears on the opposite page to
only show the parameter names.)
9781623568177_txt_print.indd 309 29/04/2013 11:51
list of parameters, each with its animation channel. Depending on
the element type, the parameters range from a few to a dozen or
more. For example, for lights, their parameters are intensity, color,
shadow darkness and shadow difusion. If an element is a camera,
the parameters include its point of interest, position, orientation,
and rotation. For shapes, the list of parameters is particularly long:
position, scale, orientation, opacity, material properties including
cast shadows, light transmission, ambient, difuse, and specular
qualities, shininess, and others.
Animating any parameter only takes a few clicks. The process
works the same regardless or the element and parameter type.
Thus, it is equally easy to change over time the position of a shape,
its color, or the intensity of a light.
In contrast to twentieth-century animation, After Efects does
not privilege movement of two-dimensional objects and characters.
Accordingly, After Efects Help denes animation as change
over timewithout specifying what can change.
The answer
is provided by the interface itself. Its design suggests that you
can animate hundreds of visual characteristics (and if we also
consider that most lters can also be animated, the list goes into
Because the software interface makes directly visible every
parameter for every object in the composition, assigning each
its own channel on the timeline, it literally invites the designer
to start animating them. You are invited to start moving and
rotating objects, changing their opacity, colors, and so on. The
same logic extends to the camera and the lights. If you add a
light to the composition, this immediately creates half a dozen
new animation channels describing lights color, position, orien-
tation, intensity, and shadow properties, each with its own
timeline channel. (Other 2D and 3D animation and layer-based
compositing software packages all share the same interface
http://help.adobe.com/en_US/AfterEffects/9.0/ (November 7, 2012).
Node-based compositing software uses a diferent interface principle: each
element in a composition is represented as node in a graph. This interface is
particularly useful for creating scenes that contain a very large number of elements
interacting with each other. Popular industry node-based compositing software
includes Fusion and Nuke.
9781623568177_txt_print.indd 310 29/04/2013 11:51
During the 1980s and 1990s the general logic of computer repre-
sentationthat is, representing everything as variables that can
have diferent valueswas systematically embedded throughout
the interfaces of media design software. As a result, although a
particular software application does not directly prescribe to its
users what they can and cannot do, the structure of the interface
strongly inuences the designers thinking. In the case of moving
image design, the result of having a timeline interface with
multiple channels all just waiting to be animated is that a designer
usually does animate them. If previous constraints in animation
technologyfrom the rst optical toys in the early nineteenth
century to the standard cel animation system in the twentieth
centuryresulted in an aesthetics of discrete and limited temporal
changes, the interfaces of computer animation software quickly led
to a new aesthetics: the continuous transformations of most (or all)
visual elements appearing in a frame.
This change in animation aesthetics deriving from the interface
design of animation software was paralleled by a change in another
eldarchitecture. In the mid-1990s, when architects started to
use software originally developed for computer animation and
special efects (rst Alias and Wavefront; later Maya and others),
the idea of animated form entered architectural thinking as well. If
2D animation/compositing software such as After Efects enables
an animator to change any parameter of a 2D object (a video clip,
a 2D shapes, type, etc.) over time, 3D computer animation allows
the same for any 3D shape. An animator can set up key frames
manually and let a computer calculate how a shape changes over
time. Alternatively, s/he can direct algorithms that will not only
modify a shape over time but can also generate new ones. (3D
computer animation tools to do this include particle systems,
physical simulation, behavioral animation, articial evolution,
L-systems, etc.) Working with 3D animation software afected
architectural imagination both metaphorically and literally. The
shapes that started to appear in projects by young architects and
architecture students in the second part of the 1990s looked as if
they were in the process of being animated, captured as they were
transforming from one state to another. The presentations of archi-
tectural projects and research begin to feature multiple variations
of the same shape generated by varying parameters in software.
Finally, in projects such as Gregg Lynns New York Port Authority
9781623568177_txt_print.indd 311 29/04/2013 11:51
Gateway (1995),
the paths of objects in an animation were
literally turned into an architectural design. Using a particle system
(a part of Wavefront animation software), which generates a cloud
of points and moves them in space to satisfy a set of constraints,
Lynn captured these movements and turned them into the curves
making up his proposed building.
Equally crucial was the exposure of architects to the new gener-
ation of modeling tools in the commercial animation software of
the 1990s. For two decades the main technique for 3D modeling
was to represent an object as a collection of at polygons. But
by the mid-1990s, the faster processing speeds of computers
and the increased size of computer memory made it practical to
ofer another technique on desktop workstationsspline-based
modeling. This new technique for representing form pushed archi-
tectural thinking away from rectangular modernist geometry and
toward the privileging of smooth and complex forms made from
continuous curves. As a result, since the second part of the 1990s,
the aesthetics of blobs has come to dominate the thinking of
many architecture students, young architects, and even already
well-established star architects such as Zaha Hadid, Eric Moss,
and UN Studio.
But this was not the only consequence of the switch from
the standard architectural tools and CAD software (such as
AutoCAD) to animation/special efects software. Traditionally,
architects created new projects on the basis of existing typology. A
church, a private house, a railroad station all had their well-known
typesthe spatial templates determining the way space was to be
organized. Similarly, when designing the details of a particular
project, an architect would select from the various standard
elements with well-known functions and forms: columns, doors,
windows, etc.
In the twentieth century mass-produced housing
only further embraced this logic, which eventually became encoded
in the interfaces of CAD software.
But when in the early 1990s, Gregg Lynn, the rm Asymptote,
Lars Spuybroek, and other young architects started to use 3D
Gregg Lynn, Animate Form (Princeton Architectural Press, 1999), pp. 10219.
I am grateful to Lars Spuybroek, the principal of Nox, for explaining to me
how the use of software for architectural design subverted traditional architectural
thinking based on typologies.
9781623568177_txt_print.indd 312 29/04/2013 11:51
software that had been created for other industriescomputer
animation, special efects, computer games, and industrial design
they found that this software came with none of the standard
architectural templates or details. In addition, if CAD software
for architects assumed that the basic building blocks of a structure
are rectangular forms, 3D animation software came without
such assumptions. Instead it ofered splined curves and smooth
surfaces and shapes constructed from these curves which were
appropriate for the creation of animated and game characters and
industrial products. (In fact, splines were originally introduced into
computer graphics in 1962 by Pierre Bzier for use in computer-
aided car design.)
As a result, rather than being understood as a composition
made up of template-driven standardized parts, a building could
now be imagined as a single continuous curved form that can vary
innitely. It could also be imagined as a number of continuous
forms interacting together. In either case, the shape of each of these
forms was not determined by any kind of a priori typology.
(In retrospect, we can think of this highly productive misuse
of 3D animation and modeling software by architects as another
case of media hybridityin particular, what I called the crossover
efect. In this case, it is a crossover between the conventions and
the tools of one design eldcharacter animation and special
efectsand the ways of thinking and knowledge of another eld,
namely, architecture.)
Relating this discussion of architecture to the main subject of
this chapterproduction of moving imageswe can see now
that by the 1990s both elds were afected by computerization
in a structurally similar way. In the case of twentieth century
commercial animation, all temporal changes inside a frame were
limited, discrete, and usually semantically driveni.e., connected
to the narrative. When an animated character moved, walked into
a frame, turned his/her head, or extended his/her arm, this was
used to advance the story.
After the switch to a software-based
production process, moving images came to feature constant
In the case of narrative animation produced in Russia, Eastern Europe and Japan,
the visual changes in a sequence were not always driven by the development of a
narrative and could serve other purposesestablishing a mood, representing the
emotional state, or simply used aesthetically for its own sake.
9781623568177_txt_print.indd 313 29/04/2013 11:51
changes on many visual dimensions that were no longer limited
by semantics. As dened by numerous motion-graphics sequences
and short lms of the 2000s, contemporary temporal visual form
constantly changes, pulsates, and mutates beyond the need to
communicate meanings and narrative. (The lms of Blake and
Murata ofer striking examples of this new aesthetics of a variable
form; many other examples can easily be found by surng websites
that showcase works by motion graphics studios and individual
A parallel process took place in architectural design. The difer-
entiations in a traditional architectural form were connected to the
need to communicate meaning and/or to fulll the architectural
program. An opening in a wall was either a window or a door; a
wall was a boundary between functionally diferent spaces. Thus,
just as in animation, the changes in the form were limited and were
driven by semantics. But today, the architectural form designed
with modeling software can change its geometry across the whole
design, and these changes no longer have to be justied by function.
The Yokohama International Port Terminal (2002), designed
by Foreign Ofce Architects, illustrates very well the aesthetics
of variable form in architecture. The building is a complex and
continuous spatial volume without a single right angle and with no
distinct boundaries that would break the form into parts or separate
it from the ground plane. Visiting the building in December 2003, I
spent four hours exploring the continuities between the exterior and
the interior spaces and enjoying the constantly changing curvature
of its surfaces. The building can be compared to a Mobius strip
except that it is much more complex, less symmetrical, and more
unpredictable. It would be more appropriate to think of it as a
whole set of such strips smoothly interlinked together.
To summarize this discussion of how the shift to software-based
representations afected the modern language of form: all constants
were substituted by variables whose values can change continu-
ously. As a result, culture went through what we can call the
continuity turn. Both the temporal visual form of motion graphics
and design cinema and the spatial form of architecture entered
the new universe of continuous change and transformation. (The
elds of product design and space design were similarly afected.)
Previously, such aesthetics of total continuity was imagined by
only a few artists. For instance, in the 1950s, architect Frederick
9781623568177_txt_print.indd 314 29/04/2013 11:51
Kiesler conceived a project titled Continuous House that, as the
name implies, is a single continuously curving spatial form uncon-
strained by the usual divisions into rooms. But when architects
started to work with 3D modeling and animation software in the
1990s, such thinking became commonplace. Similarly, the under-
standing of a moving image as a continuously changing visual form
without any cuts, which previously could be found only in a small
number of lms made by experimental lmmakers throughout the
twentieth century such as Fischingers Motion Painting (1947),
now became the norm.
Today, there are many successful short lms lasting only a few
minutes and small-scale building projects based on the aesthetics of
continuityi.e. a single continuously changing form, but the next
challenge for both motion graphics and architecture is to discover
ways to employ this aesthetics on a larger scale. How do you scale
up the idea of a single continuously changing visual or spatial form,
without any cuts (for lms) or divisions into distinct parts (for
In architecture, a number of architects have already begun
to successfully address this challenge. Examples include already
realized projects such as the Yokohama International Port Terminal,
the Kunsthaus in Graz by Peter Cook (2004), and Ordos Museum
by MAD Architects in Inner Mongolia, China (2012), as well as
those under construction, such as Zaha Hadids Performing Arts
Centre on Saadiyat Island in Abu Dhabi, United Arab Emirates.
(After the 2007 economic crisis, many ambitious building projects
in Dubai and Eastern Europe were delayed or cancelled, but China
and other countries continue to take risks and embrace the new
architectural designs made from complex continuously changing
What about motion graphics? Blake has been one of the few
artists to have systematically explored how hybrid visual language
can work in longer pieces. Sodium Fox is 14 minutes; an earlier
piece, Mod Lang (2001), is 16 minutes. The three lms that make
up Winchester Trilogy (20014) run for 21, 18, and 12 minutes.
None of these lms contain a single cut.
Sodium Fox and Winchester Trilogy use a variety of visual sources
including photography, old lm footage, drawings, animation,
type, and computer imagery. All these media are woven together
into a continuous ow. As I have already pointed out in relation
9781623568177_txt_print.indd 315 29/04/2013 11:51
The Yokohama International Port Terminal. Foreign Ofce Architects, 2002.
9781623568177_txt_print.indd 316 29/04/2013 11:51
9781623568177_txt_print.indd 317 29/04/2013 11:51
to Sodium Fox, in contrast to shorter motion-graphics pieces with
their frenzy of movement and animation, Blakes lms contain
very little animation in a traditional sense. Instead, various still
or moving images gradually fade in on top of each other. So while
each lm moves through a vast terrain of diferent visualscolor
and monochrome, completely abstract and gurative, ornamental
and representationalit is impossible to divide the lm into
temporal units. In fact, even when I tried, I could not keep track
of how the lm got from one kind of image to a very diferent one
just a couple of minutes later. And yet these changes were driven
by some kind of logic, even if my brain could not compute it while
I was watching each lm.
The hypnotic continuity of these lms can be partly explained
by the fact that all visual sources in the lms were manipulated via
graphics software. In addition, many images were slightly blurred.
As a result, regardless of the origin of the images, they all acquired
a certain visual coherence. So although the lms skillfully play on
the visual and semantic diferences between live-action footage,
drawings, photographs with animated lters on top of them,
and other media, these diferences do not create juxtaposition
or stylistic montage.
Instead, various media seem to peacefully
coexist, occupying the same space. In other words, Blakes lms
seem to suggest that deep remix is not the only possible result of
We have already discussed in detail Alan Kays concept of a
computer metamedium. According to Kays proposal made in the
1970s, we should think of the digital computer as a metamedium
containing all the diferent already existing and non-yet-invented
What does this imply for the aesthetics of digital projects?
In my view, it does not imply that the diferent media neces-
sarily fuse together, or make up a new single hybrid, or result
in multimedia, intermedia, convergence, or a totalizing
Gesamtskunstwerk. As I have argued, rather than collapsing into
a single entity, diferent media (i.e., diferent techniques, data
formats, data sources, and working methods) start interacting,
In the Compositing chapter of The Language of New Media, I have dened
stylistic montage as juxtapositions of stylistically diverse images in diferent
Kay and Goldberg, Personal Dynamic Media.
9781623568177_txt_print.indd 318 29/04/2013 11:51
producing a large number of hybrids, or new media species. In
other words, just as in biological evolution, media evolution in a
software era leads to diferentiation and increased diversitymore
species rather than less.
In the world dominated by hybrids, Blakes lms are rare in
presenting us with relatively pure media appearances. We can
either interpret this as the slowness of the art world, which is
behind the evolutionary stage of professional mediaor as a clever
strategy by Blake to separate himself from the usual frenzy and
over-stimulation of motion graphics. Or we can read his aesthetics
as an implicit statement against the popular idea of convergence.
As demonstrated by Blakes lms, while diferent media has become
compatible, this does not mean that their distinct identities have
collapsed. In Sodium Fox and Winchester Trilogy, the visual
elements in diferent media maintain their dening characteristics
and unique appearances.
Blakes lms also expand our understanding of what the
aesthetics of continuity can encompass. Diferent media elements
are continuously added on top of each other, creating the experience
of a continuous ow, which nevertheless preserves their diferences.
Danish artist Ann Lislegaard also belongs to the continuity
generation. A number of her lms involve continuous navigation
or an observation of imaginary architectural spaces. We may relate
these lms to the works of a number of twentieth-century painters
and lmmakers who were concerned with similar spatial experi-
ences: Giorgio de Chirico, Balthus, the Surrealists, Alain Resnais
(Last Year at Marienbad), Andrei Tarkovsky (Stalker). However,
the sensibility of Lislegaards lms is unmistakably that of the
early twenty-rst century. The spaces are not clashing together
as in, for instance, Last Year at Marienbad, nor are they made
uncanny by the introduction of gures and objects (a practice of
Ren Magritte and other Surrealists). Instead, like her fellow artists
Blake and Murata, Lislegaard presents us with forms that continu-
ously change before our eyes. She ofers us yet another version
of the aesthetics of continuity made possible by software such as
After Efects, which translates the general logic of computer repre-
sentationthe substitution of all constants with variablesinto
concrete interfaces and tools.
The visual changes in Ann Lislegaards Crystal World (after J.
G. Ballard) (2006) happen right in front of us, and yet they are
9781623568177_txt_print.indd 319 29/04/2013 11:51
Crystal World (after J. G. Ballard) by Ann Lislegaard, 2006. Selected stills
from a 3D animation.
9781623568177_txt_print.indd 320 29/04/2013 11:51
Crystal World (after J. G. Ballard) by Ann Lislegaard, 2006. Installation
9781623568177_txt_print.indd 321 29/04/2013 11:51
practically impossible to track. Within the space of a minute, one
space is completely transformed into something very diferent.
Crystal World creates its own hybrid aesthetics, one that
combines photorealistic spaces, completely abstract forms, and
a digitized photograph of plants. (Although I do not know
the exact software Lislegaards assistant used for this lm, it
is unmistakably some 3D computer animation package.) Since
everything is rendered in gray scale, the diferences between media
are not loudly announced. And yet they are there. It is this kind
of subtle and at the same time precisely formulated distinction
between diferent media that gives this video its unique beauty. In
contrast to twentieth-century montage, which created meaning and
efect through dramatic juxtapositions of semantics, compositions,
spaces, and diferent media, Lislegaards aesthetics ts in with
other minimalist cultural projects of the early twenty-rst century.
Today, the creators of minimal architecture and space design,
web graphics, generative animations and interactives, ambient
electronic music, and progressive fashions similarly assume that a
user is intelligent enough to make out and enjoy subtle distinctions
and continuous modulations.
Lislegaards Bellona (after Samuel R. Delany) (2005) takes the
aesthetics of continuity in a diferent direction. We are moving
through and around what appears to be a single set of spaces.
(Historically, such continuous movement through a 3D space has
its roots in the early uses of 3D computer animation: rst for ight
simulators and later in architectural walk-throughs and rst-person
shooters.) While we pass through the same spaces many times,
each time they are rendered in a diferent color scheme. The trans-
parency and reection levels also change. Lislegaard is playing a
game with the viewer: while the overall structure of the lm soon
becomes clear, it is impossible to keep track of which space we are
in at any given moment. We are never quite sure if we have already
been there and it is now simply lit diferently, or if it is a space that
we have not yet visited.
Bellona can be read as an allegory of variable form. In this
case, variability is played out as seemingly endless color schemes and
transparency settings. It does not matter how many times we have
already seen the same space, it always can appear in a new way.
To show us our world and ourselves in a new way is, of course,
one of the key goals of all modern art, regardless of the media.
9781623568177_txt_print.indd 322 29/04/2013 11:51
By substituting all constants with variables, media software insti-
tutionalizes this desire. Now everything can always change and
everything can be rendered in a new way. But, of course, simple
changes in color or variations in a spatial form are not enough to
create a new vision of the world. It takes talent to transform the
possibilities ofered by software into meaningful statements and
original experiences. Lislegaard, Blake, and Murataalong with
many other talented designers and artists working todayofer us
distinct and original visions of our world in a state of continuous
transformation and metamorphosis: visions that are fully appro-
priate for our time of rapid social, technological, and cultural
Although the discussions in this chapter did not cover all the changes
that took place during the Velvet Revolution, the magnitude of the
transformations in moving image aesthetics and communication
strategies should by now be clear. While we can name many social
factors that all could have and probably did play some rolethe
rise of branding, experience economy, youth markets, and the Web
as a global communication platform during the 1990sI believe
that these factors alone cannot account for the specic design and
visual logics which we see today in media culture. Similarly, they
cannot be explained by simply saying that contemporary consumer
society requires constant innovation, constant novel aesthetics, and
efects. This may be truebut why do we see these particular visual
languages as opposed to others, and what is the logic that drives
their evolution? I believe that to properly understand this, we need
to carefully look at media creation, editing, and design software
and its use in the production environmentwhich can range from
one person with a laptop to a number of production companies
around the world with thousands of people collaborating on the
same large-scale project such as a feature lm. In other words, we
need to use the perspective of Software Studies.
The makers of software used in media production usually do not
set out to create a revolution. On the contrary, software is created
to t into already existing production procedures, job roles, and
9781623568177_txt_print.indd 323 29/04/2013 11:51
familiar tasks. But software applications are like species within the
common ecologyin this case, a shared environment of a digital
computer. Once released, they start interacting, mutating, and
making hybrids. The Velvet Revolution can therefore be under-
stood as the period of systematic hybridization between diferent
software species originally designed to do work in diferent media.
By 1993, designers had access to a number of programs which were
already quite powerful but mostly incompatible: Illustrator for
making vector-based drawings, Photoshop for editing of continuous
tone images, Wavefront and Alias for 3D modeling and animation,
After Efects for 2D animation, and so on. By the end of the 1990s,
it became possible to use them in a single workow. A designer
could now combine operations and representational formats such
as a bitmapped still image, an image sequence, a vector drawing,
a 3D model and digital video specic to these programs within
the same design. I believe that the hybrid visual language that
we see today across moving image culture and media design in
general is largely the outcome of this new production environment.
While this language supports seemingly numerous variations as
manifested in the particular media designs, its key aesthetics feature
can be summed up in one idea: deep remixability of previously
separate media languages.
As I have stressed more than once, the result of this hybridi-
zation is not simply a mechanical sum of the previously existing
parts but new species. This applies both to the visual language
of particular designs and to the operations themselves. When a
pre-digital media operation is integrated into the overall software
production environment, it often comes to function in a new way.
I would like to conclude by analyzing in detail how this process
works in the case of a particular operationin order to emphasize
once again that media remixability is not simply about adding the
content of diferent media, or adding together their techniques and
languages. And since remix in contemporary culture is commonly
understood as comprising these kinds of additions, we may want to
use a diferent term to talk about the kinds of transformations the
example below illustrates. I called this provisionally deep remixa-
bility, but what is important is the idea and not a particular term.
(So if you have a suggestion for a better one, send me an email.)
What does it mean when we see depth-of-eld efects in motion
graphics, lms and television programs which use neither live
9781623568177_txt_print.indd 324 29/04/2013 11:51
action footage nor photorealistic 3D graphics, but have a more
stylized look? Originally an artifact of lens-based recording,
depth of eld was simulated in software in the 1980s when the
main goal of 3D computer graphics eld was to create maximum
photorealism, i.e. synthetic scenes not distinguishable from live
action cinematography. But once this technique became available,
media designers gradually realized that it can be used regardless
of how realistic or abstract the visual style isas long as there
is a suggestion of a 3D space. Typography moving in perspective
through an empty space; drawn 2D characters positioned on
diferent layers in a 3D space; a eld of animated particlesany
spatial composition can be put through the simulated depth of
The fact that this efect is simulated and removed from its
original physical media means that a designer can manipulate it
in a variety of ways. The parameters which dene what part of
the space is in focus can be independently animated, i.e. they can
be set to change over timebecause they are simply the numbers
controlling the algorithm and not something built into the optics
of a physical lens. So while simulated depth of eld maintains the
memory of the particular physical media (lens-based photo and
lm recording) from which it came, it became essentially a new
technique which functions as a character in its own right. It has
the uidity and versatility not available previously. Its connection
to the physical world is ambiguous at best. On the one hand, it
only makes sense to use depth of eld if you are constructing a 3D
space, even if it is dened in a minimal way by using only a few
or even a single depth cue, such as lines converging towards the
vanishing point or foreshortening. On the other hand, the designer
is now able to draw this efect in any way desirable. The axis
controlling depth of eld does not need to be perpendicular to the
image plane, the area in focus can be anywhere in space, it can also
quickly move around the space, etc.
In summary, when we remove depth of eld from its original
hardware home (photo and lm cameras) and move it into
software, we change where and how it can be used. We can still
use it in its original contextthat of lens-captured media. That is,
we can apply it to 3D computer graphics elements in order to make
them compatible with video or lm captured via lens. But we can
also now use it with many other media, for purely artistic efect.
9781623568177_txt_print.indd 325 29/04/2013 11:51
And we can use it in many ways that were not even conceivable
when it was part of media hardware.
Following the Velvet Revolution, the aesthetic charge of many
media designs is often derived from simpler remix opera-
tionsjuxtaposing diferent media in what can be called media
montage. However, for me the essence of this Revolution is the
more fundamental deep remixability, illustrated by this example
of how depth of eld was greatly amplied when it was simulated
in software.
Computerization virtualized practically all media creation and
modication techniques, extracting them from their particular
physical media and turning them into algorithms. This means that
in most cases, we will no longer nd any of the pre-digital techniques
in their pure original state. This is something I have already
discussed in general when we looked at the rst stage in cultural
software history, i.e. the 1960s and 1970s, examining Sutherlands
work on the rst interactive graphical editor Sketchpad, Nelsons
concepts of hypertext and hypermedia, and Kays discussions of
an electronic book (It [an electronic book] need not be treated
as a simulated paper book since this is a new medium with new
properties). We have now seen how this general idea articulated
already in the early 1960s made its way into the details of the inter-
faces and tools of applications for media design which eventually
replaced most of traditional tools: After Efects (which we analyzed
in detail), Illustrator, Photoshop, Flash, Final Cut, etc. So what is
true for depth of eld efect is also true for most other tools ofered
by media design applications.
What was a set of theoretical concepts implemented in a small
number of custom software systems accessible mostly to their
own creators in the 1960s and 1970s (such as Sketchpad or the
Xerox PARC workstation) later became a universal production
environment used today throughout all areas of the culture
industry. The ongoing interactions between the ideas coming from
the software industry and the desires of the users of their tools
(media designers, graphic designers, lm editors, and so on)led
to the further evolution of softwarefor instance, the emergence
of an new category of digital asset management systems around
the early 2000s, or the concept of production pipeline which
became important in the middle of this decade. In this chapter I
have described only one among many directions of the evolution
9781623568177_txt_print.indd 326 29/04/2013 11:51
of software applications, their tools, and media formats. As we
saw, the result of this trend was the emergence of a fundamentally
new type of aesthetics which today dominates visual and media
9781623568177_txt_print.indd 327 29/04/2013 11:51
9781623568177_txt_print.indd 328 29/04/2013 11:51
Software, hardware, and social media
In 1977 Alan Kay and Adele Goldberg imagined that a computer
would become a metamedium that would contain a wide
range of already-existing and not-yet-invented media. Exactly
as they predicted, computers have been used to invent a number
of new types of media that are not simulations of prior physical
media. The examples include navigable three-dimensional spaces
constructed with computer graphics, media databases, or sim
video games such as SimCity, The Sims, and Civilization. And,
of course, the Internet in particular has been a very productive
host for inventing new types of communication and collaboration:
email, forums, blogs, microbloging (e.g. Twitter), wiki, RSS, social
networks, etc.
At the same time, the computer metamedium was also evolving
in a second direction. While it indeed came to contain a wide
range of already-existing media, I have argued that once each
of these media was simulated in a computer their identities
changed. Sutherland, Engelbart, Nelson, Kay, and other inventors
of computer metamediums understood that the simulations of
previously existing physical or electronic media could add many
new properties to these media. As Kay and Goldberg wrote in their
article, a simulated medium can become a new medium with new
A computer breathes new life into the physical and electronic
media it simulates. Media can become dynamic (to use another
of Kays terms which he preferred to interactive). They can also
be intelligent: think of Sketchpad, which could automatically
clean up the sketches made by designers by satisfying constraints
such as parallelism. Further, they can become collectively sharable
9781623568177_txt_print.indd 329 29/04/2013 11:51
and collectively editablethink of large-scale social software
projects such as Wikipedia and OpenStreetMap. Media objects such
as pictures, sound, video, text, and code can leap from machine to
machine in a truly magical fashion: from a mobile phone to a media
player, then to a memory card, a laptop, a netbook, the Web, and
so on and so on. Imagine that you live in the sixteenth century and
you are told that you can order an image in a painting to travel by
itself and appear in another painting in another country, or that a
text in one book can lift itself and replace a text in another book?
And yet this is exactly what many of us are doing every day without
even thinking how magical this is.
I think that the scale, diversity and radicalism of these
additionsthose already invented and many more yet to be
invented in the futureis of such magnitude that exploring what
can be created with them will occupy us for a long time. And this
is one of the reasons why the digital revolution is diferent from
all previous techno-cultural revolutions. The ability to simulate
not simply one or two, but most media in a computercombined
with computer abilities to control processes in real-time, calculate,
transform inputs, test what/if scenarios, and send information
over networksopens a practically unbounded space of creative
As we followed the evolution of the computer medium from
the rst stage of invention and experimentation (1960s1970s)
to the second stage of commercialization and wide adoption
(1980s1990s), we discovered the third direction of the computer
metamedium evolutionhybridization. Translated into software,
diferent types of media started acting like species within a common
ecologyin this case, a shared software environment. Once they
were released into this environment they begin interacting,
mutating, and making hybrids. Both the simulated and new media
typestext, hypertext, still photographs, vector graphics, digital
video, 2D animation, 3D models and animation, navigable 3D
spaces, maps, location information, messages and scriptsbecame
building blocks for many new media combinations. As my examples
illustrate, such hybrids can be found at diferent scalesfrom large-
scale software systems such as Google Earth to the single images
and short motion graphics created by individual designers.
The rise of social media and social networking on the web in
the middle of the 2000s, their expansion to mobile platforms in
9781623568177_txt_print.indd 330 29/04/2013 11:51
the next few years, and the development of apps markets for these
platforms lead to new types of hybrids. The functional elements of
social and mobile mediasearch, rating, wall posting, subscription,
text messaging, instant messaging, email, voice calling, video
calling, etc.form their own media ecosystem. Like the ecosystem
of techniques for media creating, editing and navigation realized in
professional media software during the 1990s, this new ecosystem
enables further interactions of its elements. These elements are
combined in a variety of ways on diferent platforms and in
diferent social media apps. New features of social and mobile
software enter this ecosystem, and expand the pool of its elements.
The technologies behind these elements, such as multi-tier web
applications and server-side scripting, make creating new combina-
tions and new elements particularly easy in comparison to desktop
While new versions of desktop applications such as
Photoshop, InDesign or Maya that may contain new techniques
are released infrequently, online applications such as Facebook,
YouTube or Google Search can be updated by their companies very
frequently (both Google and Facebook update their code daily) and
new features can be added at any time.
The competition between leading social network services has
been one engine of new elements as well as variations on existing
elements. To give two examples from 2011, Google+ introduced
the Circles feature to enable users to organize other people using
the service into groups and to control what they post and share
with each group;
Facebook introduced a Subscribe button to allow
its users to follow public updates from particular users they like.

In the same year, the new social photo sharing service Pinterest
rose meteorically, due to its new features such as a exible layout
of photos in contrast to the xed grid layout of Facebook and
For a good explanation of how web applications work technically, see http://
Vic Gundotra, Introducing the Google+ project: Real-life sharing, rethought
for the web, googleblog.blogspot.com, June 28, 2011, http://googleblog.blogspot.
Meghan Peters, Facebook Subscribe Button: What It Means for Each Type
of User, mashable.com, September 15, 2011, http://mashable.com/2011/09/15/
9781623568177_txt_print.indd 331 29/04/2013 11:51
While this book focused on the evolution of software for media
creation and editing, this evolution is intrinsically connected with
the parallel development of hardware, including desktops, and later
laptops and mobile platforms, networks, servers, render farms,
video cards, displays, and so on. More high resolution displays,
larger and cheaper storage, faster networks, easier connectivity
between devices for media capture, storage, editing, distribution,
and playbackall these developments automatically extend the
capacities of computational media, changing what can be imagined
and designed. For example, a computer display with 35,840 x
8,000 pixel resolution (e.g. the HIPerSpace super-visualization
computer that my lab has been fortunate to use since 2008) is
not only quantitatively but also qualitatively diferent from a
display which only has 1024 x 768 pixels (the desktop standard
in the 1990s).
Similarly, the experience of using the Internet via
broadband connection is qualitatively diferent from a modem to
dial-in via analog telephone line (also the standard in the 1990s).
When Kays colleagues implemented various media editors on
their interim Dynabook in the rst part of the 1970s, most of
this software could not compete with their physical and electronic
equivalent tools. I remember my experience working on a rst
generation Macintosh in 1984; it could only show sixteen levels of
grey on a 512 x 384 pixel screen measuring 9 inches. Obviously this
was not sufcient for me to immediately drop my oil brushes and
paints and switch to computers. So, in some sense, the rst period
of media computerizationbetween the completion of Sketchpad
in 1963 and the release of PageMaker in 1985was theoretical.
During this period, the conceptual principles and the key algorithms
necessary for detailed simulation of physical media were developed,
before the sufciently cheap hardware was available. For instance,
during the 1960s many computer scientists learned about Sketchpad
by reading Sutherlands PhD thesis since the machine on which it
runthe TX-2 computerexisted only at MIT. (This is another
interesting characteristic of computer media revolutionit was
theorized in detail before it occurred in practice.)
Photographs of my labs work with the HIPerSpace super-visualization computer
are available at http://lab.softwarestudies.com/2008/12/cultural-analytics-
hiperspace-and.html. The HIPerSpace computer is described at http://vis.ucsd.edu/
9781623568177_txt_print.indd 332 29/04/2013 11:51
But during the second part of the 1990s PC hardware become
advanced enough to run simulations of most media at a sufcient
delity that was comparable with the professional standards
already in place. This included software for graphic design, CAD,
3D modeling, 2D and 3D animation, print layout, digital photo
manipulation, and audio and video non-linear editing. In many
cases, these applications were now responding to users actions
sufciently fast to approach the level of interactivity conceptualized
by Kay and Goldberg in 1977 (although 3D animation and video
compositing even today may require long rendering times). As a
result, simulated media became truly useful, and also accessible, to
large numbers of people outside of computer laboratories and large
media companies. Within half a decade, most culture professionals
abandoned physical media for their simulated equivalents.
When I visited well-known electronic musician, writer and
artist DJ Spooky in his Tribeca apartment in New York in January
2005, I did not nd any musical instruments, traditional or
electronic. The only instrument Paul Miller (aka DJ Spooky
That Subliminal Kid) owned was his 15-inch PowerBook laptop,
made by Apple. This was his Dynabook: a self-contained
knowledge manipulator in a portable package the size and shape of
an ordinary notebook.
Although this Dynabook did not have
Smalltalk, it ran another programming environment which was
powerful, fast and allowed for visual programmingMAX, the
language of choice worldwide for tens of thousands of electronic
musicians, VJs, dancers, theatre performers and others working
with diferent forms of real-time performance.
While the evolution of hardware enabled dissemination of
media software to professional communities in the 1990s, in
the 2000s this evolution also enabled the new stage in software
design and usesocial media (2004). The few companies which
dominated the eld of professional media applications (Adobe,
Apple, Autodesk) were joined by a multitude of new companies
and countless start-ups focused on developing tools and services
for the Web and mobile platforms.
The new software categories include social networking (e.g.
Facebook), micro-content services (e.g. Twitter), media sharing
Kay and Goldberg, Personal Dynamic Media, p. 394.
9781623568177_txt_print.indd 333 29/04/2013 11:51
websites (YouTube, Vimeo, Picasa, Flickr, etc.); consumer-level
software for media organization and light editing (e.g., iPhoto);
blog editors (Blogger, WordPress); and many others. Keep in
mind that softwareespecially web and mobile applications and
services designed for consumerscontinuously evolves, so some
of the categories above, their popularity, and the features of
particular applications and social network services may change by
the time you are are reading this. One graphic example is the shift
in the identity of Facebook. During 2007, it moved from being
yet another social media application that was competing with
MySpace to a social OS that aimed to combine the functionality
of previously diferent applications in one placereplacing, for
instance, stand-alone email software for many users.)
None of the software apps and websites of the social media era
function in isolation. Instead, they participate in the larger ecology,
which includes search engines, recommendation engines, blogging
systems, RSS feeds, and other web technologies; inexpensive
consumer electronic devices for capturing and accessing media
(digital cameras, mobile phones, music players, video players,
digital photo frames, internet enabled TVs); and technologies that
enable transfer of media between devices, people, and the web
(storage devices, wireless technologies such as Wi-Fi and WiMax,
and communication standards such as USB and 4G). Without
this ecology most web services and mobile apps would not be
possible. Therefore, this ecology needs to be taken into account
in any discussion of social networks and their softwareas well
as consumer-level content access and media development software
designed to work with web-based media sharing sites. And while
the particular elements and their relationship in this ecology are
likely to change over timefor instance, almost all media content
may eventually be available over computer networks; communi-
cation between devices may similarly become fully transparent;
and the very rigid physical separation between people, devices they
control, and non-smart passive space may become blurredthe
very idea of a technological ecology consisting of a number of
interacting parts which include software is not unlikely to go away
any time soon. Thus, if one day I were to write a detailed account
of social media, I would need to discuss consumer electronics,
network architectures and protocols and other elements of this
ecology as much as the social software itself.
9781623568177_txt_print.indd 334 29/04/2013 11:51
Media after software
Any summary of a 100,000-word book of theoretical arguments
cant cover all important points. However, I will risk this because I
know that even an incomplete summary will still be useful for the
readers. Here are some of the proposals developed in this book
about the experiences and meanings of media for contemporary
designers who create it using software applications, and for the
users of interactive media applications and services:
1 The computer is not a new mediumit is the rst
metamedium: a combination of existing, new, and yet to
be invented media. (This is the argument of Kay, which I
take as my starting point.)
2 A medium (as it exists in software) is a combination
of particular techniques for generation, editing and
accessing content. (I use the generic term access as an
alias to the longer list of termsnavigating, browsing,
viewing, listening, reading, and interacting.) Softwarization
virtualizes already existing techniques and adds many new
ones. All these techniques together from the computer
metamedium. Any single medium uses a subset of these.
New techniques and their variations are constantly
being developed which changes the identity of each
medium that uses them. For the users of popular
commercial media software, a medium changes with each
software release.
3 What we identify by a conceptual inertia as properties
of diferent mediums are actually the properties of media
software their interfaces, the tools, and the techniques
they make possible for accessing, navigating, creating,
modifying, publishing and sharing media documents.
4 Following the same logic, properties of any media object
are no longer fully dened by the content and formats of
the les storing the information. They now also depend
on the software used to access this object. For example,
depending on whether the same image is accessed via a
default media viewer, a consumer app for media access and
9781623568177_txt_print.indd 335 29/04/2013 11:51
editing, or professional editing software such as Photoshop,
its properties change signicantly.
5 The techniques that make up the computer metamedium
can be divided into two categories. General-purpose (i.e.
media-independent) techniques are implemented to work
in the similar way on all media types (for example: select,
copy, search, lter, etc.) Media-specic techniques can only
work on particular data structures (for example, you can
increase amplitude of a sound track or reduce the number
of vertices in a 3D shape, but not vice versa). Each software
medium combines some media-independent and some
media-specic techniques.
6 The idea of a data structure leads us to an alternative
denition of a software medium. As dened by application
software and experienced by users, a medium is a pairing
of a particular data structure and the algorithms for creation,
editing and viewing the content stored in this structure.
From this perspective, each of the categories of media
development software can be said to dene its own
mediumbecause the programs ofered in each of
these categories typically (but not always) share the same
fundamental data structure. The examples that t this are
vector graphics editors, raster graphics editors, 2D animation
and motion graphics software, 3D computer graphics
software, sound editors, text processors, and HTML editors.
7 Following the rst stage of the computer metamedium
invention, we enter the next stage of media hybridity and
deep remixability. The unique properties and techniques
of diferent media became software elements that can be
combined together in previously impossible ways. Both
the simulated and new media typestext, hypertext, still
photographs, digital video, 2D animation, 3D animation,
navigable 3D spaces, maps, location informationcame
to function as building blocks for many new media
This condition is not the simple consequence of the
universal digital code used for all media types. Instead, it
is the result of the gradual development of interoperability
9781623568177_txt_print.indd 336 29/04/2013 11:51
technologies including standard media le formats, import/
export functions in applications, and network protocols.
8 The previous formulations lead us to view contemporary
media development using a model of biological evolution
and its concept of massive numbers of species that share
common traitsaway from the modern model of a small
number of very diferent mediums with their unique
languages. Instead of trying to place any particular project,
app, or web service in some category selected from a small
number, we can instead view it as a combination of the
techniques selected from a very large pool. Some of these
combinations occur more often; others may only occur
once. The successful combinations become popular, leading
to similar projects; and some may become design patterns
used in numerous applications.
Software epistemology
One of the key ideas developed in this book is that the computer
metamedium is characterized by permanent extendibility. New
algorithms and techniques that work with common media data
types and le formats can be invented at any time by anyone with
the right skills. These inventions can be distributed instantly over
the web, without a need for the large resources that were required
in the twentieth century to introduce a new commercial media
device with new functions or a new media format. Use of open
source and free software licenses and web-based hosting services
repositories such as GitHub encourages people to collectively
expand existing software tools, which often leads to their rapid
The permanent extendibility of the computer metamedium has
important consequences not only for how we create and interact with
media, but also for the techniques of knowledge in a computerized
society (Knowledge in Computerized Societies is the subtitle of
the opening section of the celebrated 1979 book The Postmodern
Condition: A Report on Knowledge by Jean-Franois Lyotard).
Turning everything into data, and using algorithms to analyze it
changes what it means to know something. It creates new strategies
9781623568177_txt_print.indd 337 29/04/2013 11:51
that together make up software epistemology. Epistemology is
a branch of philosophy that asks questions such as what is
knowledge, how it can be acquired, and to what extent a subject
can be known. Digital code, data visualization, GIS, information
retrieval, machine learning techniques, constantly increasing speed
of processors and decreasing cost of storage, big data analytics
technologies, social media, and other parts of the modern techno-
social universe introduce new ways of acquiring knowledge, and in
the process redene what knowledge is.
For instance, it is always possible to invent new algorithms (or
new ways to scale existing algorithms to analyze big data faster)
that can analyze the already existing data in ways the previous
algorithms could not. As a result, you can extract additional
patterns and generate new information from the older, already
analyzed data.
Algorithms and software applications that analyze images and
video provide particularly striking examples of this capacity to
generate additional information from the data years or even
decades after it was recorded.
In Blowup, a 1966 lm directed by Michelangelo Antonioni, the
photographer who takes pictures of a couple kissing in the park
uses the strategy of printing progressive enlargements of one area
of the photograph, until a grainy close-up reveals a body in the
grass and a man with a gun hiding in the trees.
During the time that this lm was being created and unknown to
its director, computer science researchers were already developing
the new eld of digital image processing, including algorithms
for image enhancement such as sharpening of edges, increasing
contrast, and reducing noise and blur. The early articles in the eld
show the blurry photographs taken by surveillance planes, which
were sharpened by the algorithms. As I already explained earlier,
today many of these techniques are built into all image manipu-
lation software such as Photoshop, as well as in the rmware of
digital cameras. They became essential to both consumer photog-
raphy and commercial visual media, as every published photograph
in mass media rst goes through some software adjustments.
Contemporary DSLR models and high-end compact digital
cameras can record images both in JPEG and Raw formats. With
JPEG format, an image is compressed, which limits the possibilities
for later extraction of additional information using software.
9781623568177_txt_print.indd 338 29/04/2013 11:51
Raw format stores the unprocessed data from the cameras image
sensor. The use of this format assumes that a photographer will
later manipulate the photo in software to get the best results from
the millions of pixels recorded by the camera. In his guide for the
use of the two formats by photographers, William Porter explains:
Working with a Raw le, youll have a greater chance of recov-
ering highlights in overexposed clouds, rescuing detail from areas
of shadow or darkness, and correcting white balance problems.
Youll be able to minimize your images noise and maximize their
sharpness, with much greater nesse.

This means that new software with better algorithms can
generate new information from a photo in Raw format captured
years earlier. (See the examples from Porters article available online
for a dramatic demonstration of this.
) In another example of
software epistemology, in the demo presented at one SIGGRAPH
annual conferences, a few lm shots of 1940s Hollywood lm
were manipulated in software to re-generate the same shots from a
diferent point of view.
In visual efects production today, one of the most widely used
operations is the algorithmic extraction of the position of the video
camera that was used to capture a shot. This operation is called
motion tracking and it exemplies how information that is not
directly available in the data can be inferred by algorithms. (The
extracted camera position is used to insert computer graphics into
the live action footage in the correct perspective.)
Another important type of software epistemology is data
fusionusing data from diferent sources to create new knowledge
that is not explicitly contained in any of them. For example,
using the web sources, it is possible to create a comprehensive
description of an individual by combining pieces of information
from his/her various social media proles and making deductions
from them.
Combining separate media sources can also give additional
meanings to each of the sources. Consider the technique of the
automatic stitching of a number of separate photos into a single
panorama, available in most digital cameras. Strictly speaking, the
William Porter, Raw vs. JPEG: Which should you shoot and when?, techhive.
com (September 19, 2012).
9781623568177_txt_print.indd 339 29/04/2013 11:51
underlying algorithms do not add any new information to each of
the images (i.e., their pixels are not modied). But since each image
now is a part of the larger panorama, its meaning for a human
observer changes.
The abilities to generate new information from the old data, fuse
separate information sources together, and create new knowledge
from old analog sources are just some techniques of software
epistemology. In my future publications I am hoping to gradually
describe other techniquesas I am teaching myself data mining and
other algorithmic knowledge techniques of knowledge commonly
used by contemporary software societies.
In the beginning of the book I asked: what is media after
software? If artistic mediums were traditionally dened by the
techniques and representational capacities of particular tools and
machines (brushes, ink, paper, musical instruments, a printing
press, a photo camera, a lm camera, video cameras and editing
equipment), what happens to this concept after most of these tools
and techniques are simulated in a single software environment? In
other words, what is a medium as dened by software applica-
tions used to create, edit, distribute, and access it?
While media efects, media representation, media
industry, media theory and media history have been exten-
sively discussed in large number of books and articles in a number
of academic disciplines, this literature does not usually include
analysis of software tools and platforms. In contrast, the vast
universe of how to books, instructional videos and tutorials
contains very little theory, because the goal of all these publica-
tions is practical instruction. (My search for Photoshop in
amazon.com under books on August 12, 2013 returned 9,405
publications; the search for After Efects returned 1,201 results,
and the search for 3ds Max returned 1,972 results.) The aim of
my book was to help bridge the gap between these two separate
universes of theory and practice.
Following the question of what it means to create media with
software took us on a long journey through a few decades. We did
nd some possible answers that I hope you have found interesting
and provocative. But of course, since both this book and all my
writing are directed rst of all to media practitionersprofes-
sionals and students creating new software applications and tools,
graphic designs, web designs, motion graphics, animations, lms,
9781623568177_txt_print.indd 340 29/04/2013 11:51
space designs, architecture, objects, devices, and digital artyou
can do more than simply agree or disagree with my analysis. By
inventing new techniques, or through the innovative application
of existing techniquesand by nding new ways to represent the
world, the human being, and the data, and new ways for people to
connect, share, and collaborateyou can expand the boundaries of
media after software.
9781623568177_txt_print.indd 341 29/04/2013 11:51
9781623568177_txt_print.indd 342 29/04/2013 11:51
A to D see analog to digital
Abel, Robert 288
Abyss, The 21, 294
ACM Digital Library 389
Acrobat 38, 269
8.0 188
designers 189
Acrobat Reader 195, 225
Acrobat User Interface 188
ActionScript 16
Adobe 27, 46, 93, 113, 147, 211,
283, 302, 333
website 304
Adobe Acrobat see Acrobat
Adobe Bridge 144
Adobe Creative Suite 302
Adobe Photoshop Touch 26
Adobe Premiere Pro 304
Advanced Research Projects
Agency Network 1834
aerial imagery 134
aesthetic theory 149, 150
aesthetics 4, 45, 120, 1212, 182,
236, 244, 252, 258, 259,
277, 297, 303, 311, 318,
319, 322, 323, 324, 325
Baroque 215
hybridity 25467
psychedelic 298
After Efects 2, 24, 39, 41, 44, 45,
46, 49, 58, 75, 84, 90, 124,
144, 179, 205, 246, 247,
248, 249, 255, 261, 278,
2829, 291, 293, 296, 300,
302, 305, 306, 307, 310,
311, 319, 324, 325
4.0 47
After the Great Divide 250
Agre, Phil 12
Ajax techniques 218
Aldus 283
Aldus Pagemaker 41
Alexander, Amy 1314
algorithms 114, 128, 136, 137,
1389, 146, 179, 181, 182,
184, 197, 199204, 215,
216, 219, 220, 222, 270,
325, 337, 338, 340
Algorithms Plus Data Structure
Equals Programs 207
Alias 41, 44, 148, 290, 311, 324
AliasWavefront 148
Alice in Wonderland 259
Alien 294
Alsace 185, 187, 191, 196, 197,
Amazon 7, 9, 35, 92, 190, 191
Amazon Kindle 109
amplication 3237
analog to digital 153
Anderson, Peter 299, 303
Android 7, 24, 26, 29, 182, 228
animation 29, 129, 145, 170
Anime Music Video 46
Antonioni, Michelangelo 338
Apeloig, Philippe 300, 303
9781623568177_txt_print.indd 343 29/04/2013 11:51
Aperture 24, 124, 144, 225
API 190, 191, 237
App Store 108
Apple 6, 7, 39, 47, 57, 84, 85, 93,
105, 107, 108, 109, 147,
164, 225, 333
statistics 108
Apple IIe 20
AppleScript 211
apps 9, 17, 24, 25, 31, 47, 50,
108, 124, 148
architecture 41
Architecture Machine Group 161
Arns, Inke 14
ARPANET see Advanced
Research Projects Agency
art 162
Art+Com 170, 187, 192
Artstor 228
artworks 139
ASCII text 209
Aspen Movie Map 80, 161, 164,
165, 176, 181
Asympote 312
Atari 84
Atkinson, Bill 105
authoring 163
AutoCAD 41, 44, 75, 312
Autodesk 27, 85, 283, 333
automation 128, 129
Avatar 259
Avid 59, 283
Bach, J. S. 118
Balthus 319
bandwidth 134
Bangalore 32
Barthes, Roland 81
Bass, Saul 40, 46, 279, 288
Your Painting project 227, 228
Beach Boys 145
Beatles 145
Bell, Alexander Graham 155
Bell, Daniel 100
Bell Laboratories 57, 85
Bellona 322
Berners-Lee, Tim 21, 161
Bzier, Pierre 313
Bilal, Enki 259
Bing Maps 143, 196
biological evolution 1678, 177,
theories 168
bitmap image 209
bit-mapped display 99
Blake, Jeremy 258, 2601, 307,
314, 315, 319, 323
Blender 24, 39, 203
Blogger 2, 24, 27, 47, 206, 334
blogs 1
Blowup 338
Blue Brain Project 236
Bogost, Ian 15, 42
Bolter, Jay 15, 58, 59, 61
books 78
bootstrapping 83
Bordwell, David 276
Borges, Jorge 11
brands, global 6
Breathless 81
browsers 35, 50, 150, 193
Bruges, Jason 164
Brunelleschi, Filippo 40
Bruner, Jerome 98, 233
theory of multiple mentalities
98, 100
Burton, Tim 259
Bush, Vannevar 11, 63, 83
Memex 73
C++ 238
CAD see Computer Aided Design
Cameron, James 21, 58, 259
Cars 280
9781623568177_txt_print.indd 344 29/04/2013 11:51
Cartesian space 294
Castells, Manuel 15, 16
CD-ROMs 166
cel animation 145
CGI 259, 290, 291, 295
chat 2
children 98, 102, 104
Chiroco, Giorgio de 319
Chrome 2, 24
Chun, Wendy 15
cinema 170, 236
cinematography 193, 199, 274
CinePaint 206, 247
Cisco Systems 7
Civilization 329
website 6
code studies 42
Codecademy 17
coding 133
cognition 123
cognitive psychology 232
Coldplay 257
color reproduction 299
Combustion 246
Coming of Post-Industrial Society,
The 100
commands 114, 212, 225, 297
Common 257
communication 205
communication studies 34, 228
communication theory 35
compositing 27782
Computational Culture 12
computational media 91, 93, 94,
96, 97, 100, 135
Computer Aided Design (CAD)
93, 230, 283, 312, 313, 333
computer drawing 93
computer graphics 89, 118
Computer History museum 40
computer industry 55
computer metamedium 45
computer painting 93
computer programming 208, 220,
223, 239
computer programs 34
computer science 10
computer scientists 31, 84, 91,
constructivism 121, 122
contrast 299
Convert 257, 268
Cook, Peter 315
Cox, Geof 14
crafts 2
Cramer, Florian 14, 15
creativity 97
Crystal World 319, 322
Cubism 83
cultural anthropologists 32
cultural heritage
digitization 226
cultural memory 183
cultural studies 205
cursor 128
cut and paste 114, 122
cyberculture 75
cyberspace 75
cyborgs 75
D to A see digital to analog
dada sonication 115
Darwin, Charles 1767, 239
On the Origin of Species 176
literary 33
data manipulation 11012
data structures 197, 207, 201,
databases 389, 143
Datapoint Corporation 85
DC Comics 180
De Stijl 82
aesthetics 121
de.licio.us 84
9781623568177_txt_print.indd 345 29/04/2013 11:51
deep remix 45, 270, 318
deep remixability 305, 324, 325
Derrida, Jacques 81
visual 45
design studies 205
designers 93
desktops 60
Diesel 164
Digg 84
digital art 139, 166
Digital Art 162
digital compositing 1447
digital computers 59, 135
digital culture 25
Digital Darkroom 203
Digital Efects 250
digital imaging 201
digital libraries 389
digital media 59, 132, 152, 155,
scholars 81
digital photography 612
digital to analog 153
digital video recorder (DVR) 35
Director 21, 44
Disney 85, 266
DJ Spooky 333
DJ/VJ/live cinema performances
Dourish, Paul 12
Dow Jones Industrial Average 67
Dreamweaver 24, 39, 124, 165,
206, 247, 302
DSLR 338
Duf, Thomas 145, 146
Duncan, Isadora 40
DVD players 153
DVR see digital video recorder
Dynabook 6491, 109, 231, 333
Earl King 80
Earth Viewer 192
eBay 7, 9
Edison, Thomas 154, 155, 156
editing 163
Eisenstein, Sergei 40
electronic devices 108
electronic literature 42
electronics industries 223
email 2
encoding 154
Engelbart, Douglas 5, 13, 40, 56,
60, 63, 71, 72, 73, 75, 78,
83, 91, 92, 93, 96, 97, 102,
104, 135, 162, 184, 225,
epistemology 33741
EPS 30
Epstein, Jean 119
Equalize 132
Ethernet 57
Europeana 227, 228
evolution 45, 168
evolutionary biology 237, 239
evolutionary theory 177
Excel 30, 220
exegesis 78
Expanded Cinema 92
export 296307
Expressionism 83
Expressive Processing 103
Extrude 130
Facebook 1, 2, 6, 9, 24, 27, 30,
39, 47, 84, 124, 148, 149,
fair trade certication 37
Fall Joint Computer Conference
Farm Security Administration 227
feature lms 179
Ferro, Pablo 46, 254, 279, 298
Fessenden, Reginald 155
FFmpeg 211
Field Studies 185
9781623568177_txt_print.indd 346 29/04/2013 11:51
le formats 21519
les 135
lm and TV studies 205
Final Cut 2, 24, 32, 39, 75, 118,
124, 144, 179, 206, 247,
283, 325
Final Cut Pro 47
ndability 115, 119, 122, 124,
ne art 2
Finlay, Karen 164
Firefox 2, 24, 47, 92
Fischinger, Oskar 266, 2867,
298, 315
Flame 39, 206, 246, 252, 294
Flash 31, 84, 179, 206, 246, 300,
302, 325
Flickr 30, 38, 47, 84, 92, 148,
150, 184, 191, 1967, 227,
Flickrvision 3D 189
folders 29, 101, 135
Ford 7
Foreign Ofce Architects 314
Foucault, Michel 81
4G 334
Foursquare 38
Frame Imagery 279
Frantz, Matt 248
Freehand 58
Freud, Sigmund
Interpretation of Dreams 169
Fry, Ben 105
Fujihata, Masaki 185, 187
Fuller, Matthew 11, 14, 15
functionality 221
functions 137, 220, 221
Fusion 247
Futurism 82, 83
futurists 236
Galloway, Alex 15
game platforms 42
game studies 205
GarageBand 145
Gardner, Howard 233
General Electric 7
General Motors 7
General Perspective Projection 36
Geocommons 38
Geographical Information Systems
(GIS) 456, 338
Gesamtkunstwerk 120, 318
Giedion, Sigfried 5
Gimp 39, 50, 205, 247
GIS see Geographical Information
Glass, Philip 118
global culture industry 27
global information society 8
globalization 8, 85
Gmail 24
GMC Denali 257
Go 268, 271, 293
Godard, Jean-Luc 279
Goldberg, Adele 64, 70, 97, 99,
100, 102, 104, 105, 110,
123, 140, 141, 150, 161,
163, 176, 18081, 329,
Goodman, Nelson 65
Languages of Art 65
Google 1, 6, 7, 9, 23, 24, 27, 39,
148, 149, 151, 190, 191,
198, 226, 287
algorithms 71
SketchUp 203
Google Analytics 30
Google Books 123
Google Docs 25, 220
Google Earth 2, 36, 37, 38, 45,
48, 84, 143, 148, 163, 165,
171, 182, 187, 192, 193,
195, 196, 204, 218, 234,
237, 277
5.0 192, 195
9781623568177_txt_print.indd 347 29/04/2013 11:51
6.0 193
GitHub 149, 330, 337
Street View 36, 164, 195
Google Image Search 114
Google Maps 16, 24, 37, 38, 47,
92, 143, 184, 190, 193
Google Play 17, 26, 35
Google Scholar 147
Google Search 47, 331
Google Trends 50, 118
Google+ 331
Goriunova, Olga 14
GPS 182, 184, 185, 192
graphic design 170
Graphic Design for the 21st
Century 299
Graphical User Interface (GUI)
20, 21, 55, 57, 62, 97, 99,
100, 101, 102, 108, 148,
188, 189, 211, 212, 221,
222, 233, 237, 283
applications 109
graphics card 91
Graz, Kunsthaus 315
Greenberg, Clement 120, 236
Greenberg, Richard 288
Greenberg, Robert 288
Greenpeace data 37
Grifth, D. W. 40
Grusin, Richard 58, 59, 61
GUI see Graphical User Interface
Gutenberg, Johannes 40
Gutenberg Bible 94
hackers 91, 93
Hadid, Zaha 248, 300, 303, 312,
Hall, Stuart 35
Hamilton, Richard 251
hard drives 32
hardware 92, 96, 332, 333
Harrell, Fox 12
Harry 251
Hayles, Katherine 14, 1516
designers 100
experts 100
HD resolution 154
HDTV 302
Heareld, John 278
Hewlett-Packard 84
High-pass 132
HIPerSpace 332
Hiroshima Mon Amour 81
Hitchcock, Alfred 279
Hoch, Hannah 278
Hockney, David 251
hosting websites 28
HTML 42, 84, 170, 179, 201,
209, 218
markup 31
HTML5 80
Hugo 259
Human Connectome Project 236
human perception 123
humanists 32
Huyssen, Andreas 250
hybrid media 166, 171
hybrid visual language 258
hybridity 16176, 18495, 204,
257, 267
aesthetics 25467
hybridization 176, 200, 247, 324,
HyperCard 21, 84, 105
hyperlm 7980
hyperlinking 71, 151, 1789
hypermedia 71, 79, 161, 274, 325
hypermodernity 275
hyperrealism 260
hypertext 63, 71, 73, 75, 78, 79,
80, 81, 82, 325
IBM 20, 135
Watson R